Empowering Patients and Their Families to Solve Rare Disease Mysteries: Dr. Lukas Lange, CEO & Co-Founder of Probably Genetic

Hillary Acer: Hi, I'm Hillary Acer welcoming you to Raise the Line with Osmosis from Elsevier, an ongoing exploration about how to improve health and healthcare. As we've discussed with other guests who have joined us on the podcast as part of our core focus on rare diseases, one major problem in the space is finding enough patients with any particular rare condition to conduct clinical trials. Our guest today Dr. Lukas Lange cofounded a company called Probably Genetic, which is seeking to address that problem through free direct-to-consumer genetic testing and we're looking forward to finding out how it all works. Before launching Probably Genetic in 2019, Dr. Lange was a biotech advisor for the Bill and Melinda Gates Foundation. He also holds a PhD in bioinformatics and genetics from the University of Oxford. Thank you so much, Dr. Lange, for being with us today. Thank

Dr. Lukas Lange: You so much for having me. I'm a huge Osmosis fan, and have read many, many of the articles you've put out. So I'm excited to be here. 

Hillary: We're really happy to hear that and we'll probably touch a little bit more on your rare disease education in a moment, but let's start by hearing more about what first got you interested in science overall?

Dr. Lange: That's a pretty broad question. I think that honestly, the very first thing that got me excited about scientific progress was the development of hydrogen powered cars in high school. I remember I read an article when I was maybe you know, thirteen or fourteen and it was fascinating. Based on that I got more into it. I thought the vision of having cars where you've just got water coming out of the exhausts was super exciting.

Hillary: Wow, that is a very interesting kind of first inspiration, that somehow led you down the path of getting into the rare disease, space and genetic space. So how did you make the jump from cars to genetics?

Dr. Lange: Yeah, so based on that hydrogen motivation, I studied chemical engineering. I did that in Munich and got an undergraduate degree and ended up working not on hydrogen powered cars, but on electric car batteries, which at the time, felt like a nascent, you know, very juvenile field, and now is obviously how the car industry is run. Then I went to the University of Cambridge for my master's degree and ended up in a group that also focused on electrochemistry, but they focused on bio diagnostic devices as an application, so building blood glucose sensors.

So, that was really my first foray into healthcare, and then decided I wanted to learn more about healthcare broadly, but specifically genetics, because I just got excited about genetics as a field, and then won a Rhodes Scholarship and went to Oxford. I had funding to fund a PhD with that scholarship, and wanted to study genetics and serendipitously landed in a rare genetic disease group that just so happened to not only work on rare diseases as a priority, but one of my PhD supervisors actually set up the 100,000 Genomes Project in the UK and raised all the money from the British government to pull off that amazing project, and that's sort of how I got thrust into it, and then got exposed to all the challenges rare disease families have to deal with, and that's kind of how it all started.

Hillary: Wow, we've covered rare diseases at length across the podcast, but I don't think we've touched as much on this project in the UK. Can you tell us a little bit more about the 100,000 Genomes Project? 

Dr. Lange: Yes, it's a fascinating study. To my knowledge, it's still the largest rare disease study on the planet. The original motivation, I would say was really twofold. The first one was to sequence 100,000 genomes of people who predominantly have rare diseases, mostly non-cancer, but also a lot of cancer patients, to one, figure out if we could deliver diagnoses for them, and two, to understand if whole genome sequencing as a technology should really be the next wave of how we do diagnostics in rare diseases. At the time, and sort of still, today it’s an open question. From my perspective, as a scientist, it makes perfect sense, but I understand there's cost considerations around that, too. So, that was one part. 

I think the other part that the UK really got, right, that I think other countries still struggle with, is that not only can you do tremendous things for patients and for healthcare doing that, but you can also build up this phenomenally valuable resource. For a country like the UK where if you think of healthcare data as a resource, similar to oil or coal -- or whatever else you have as an export -- then you can really do fantastic things with that and build a whole new industry and I think they accomplished both things with it.

Hillary: Fascinating, and that somehow led you down the path of founding a company, Probably Genetic. So, tell us about that mission and what you're up to lately.

Dr. Lange: So, when I was doing a PhD, I mostly worked on bioinformatics, meaning patients would already spend their seven to eight years on average being undiagnosed, and then miraculously, they would get “lucky” enough to see a physician who somehow knew about the 100,000 Genomes Project and who would take a DNA sample for them that would go to our lab. The sample at the lab would get sequenced, and then once we have the raw genetic data, somebody like me would start working on that data to try to figure out, why is this patient sick, when somebody else is healthy? We'd analyze their DNA and try to figure it out. For a good chunk of my PhD, I was really just focused on that numbers side of the problem…like the technology to analyze DNA. 

But once you realize that all the patients we see are the lucky ones and being lucky already means, you know, seven plus years of not knowing what the hell is going on with them, that really makes a question of how the entire system works. Especially since this is not a genetic diagnostic problem, right? Like we have the technology to get these patients tested and figure out what's going on with a really good chunk of them -- not for all of them -- but for a lot of the diseases, we know how the mechanisms work. So, the question that ensued for me was, well, what's going wrong in the system that we still have millions of people -- we think about 200 million people globally -- who just never make it to the right test?

That's really where sort of the finding motivation for Probably Genetic was born, which is, let's try to figure out how we can actually get those 200 million people who are undiagnosed, diagnosed. 

The question that we had to answer is, well, if that's what our mission is, how the hell do we pull it off? The first place we started in was at the time pretty novel, but now feels really commonplace, which is that the symptomatic data on all of these patients lives in electronic health records. We thought if we can just figure out ways to analyze the symptomatic data and tease out patterns and say, “Hey, I don't think this is just any kid, I think this is Rett syndrome patient. We need to get them tested,” then we could find those patients. 

So, my co-founder and I started developing this algorithm in our spare time that tried to plug in symptomatic data on patients that we got out of EHRs and see if we can we get them diagnosed. Then we made the typical mistake that all techies do, which is we did not talk to patients, we did not talk to doctors, we did not talk to the hospital. We just spent six months in our “magic time” as in, like, between midnight and twelve and on weekends developing this algorithm and then we went to the hospital and told a PI there, “Hey, we've got this amazing algorithm. Can we get access to the EHR and show you how it works.” And they said “Amazing. Which one of our EHRs do you want to get access to?” 

Then we started realizing just how difficult this problem is, because that particular hospital didn't have one EHR, they've got four, because cardiology has one, urology has one, oncology has one and then there's a general one for the whole hospital. They're all different, and most of them don't have structured data in them. So very quickly we realized the problem is not actually the algorithm, which is what we were excited about, it's everything else. And if you then think about the problem we're trying to solve -- which is we're trying to find a needle in the haystack patient -- you cannot possibly scale that across the tens of thousands of hospitals on the planet. 

Then at the same time, I went to the American Society for Human Genetics meeting in San Diego in 2018, and met this mom of a young girl called Mila who later passed away from Batten Disease CLN2. At the time, Mila had already been diagnosed and I was telling her mom, Julia Vitarello, about this frustration we had and how we thought we could solve the problem, but couldn't. Julia told me, “I know exactly what you're talking about. Before Mila got diagnosed, I knew that she had an autism diagnosis, I knew that she had seizures, I knew that she could climb stairs previously and couldn't any more, and I knew that she was losing her vision.” In my head, I thought ‘this mom knew the kids phenotype long before the kid got diagnosed.’ Maybe the way to do this is just to give the parents tools to phenotype the kids, not perfectly, but just good enough. And if we can get the phenotypic data from the parents, we should then be able to identify patients who should get testing done. So, that's sort of where the journey began.

Hillary: Incredible. I mean, I can hear how passionate you are about this and I'm just grateful for people like you who haven't given up when they face difficult challenges, like a hospital with four different EHR systems. Clearly, this mother, Julia, has been a great inspiration, but also will probably benefit in the future. So, I'd love to hear a little bit more about the tests themselves and how they actually work, detecting various genetic diseases, and maybe just the technology behind that. How does it work?

Dr. Lange: Sure, I'll just explain the process pretty quickly. So, I mentioned where the insight came from, right? The parents and often the patients, depending on the disease, they can describe their own symptoms and so the idea was, what if the way to find these patients is not to integrate with healthcare records in hospitals; what if we just have to build technology that we can put on the web? So, we decided we're going to build phenotyping technology that we put on a website and then the idea is if you’re a mom of a kid that's got all of these crazy symptoms and you're really worried about them and nobody's helping you and you are in between running errands at Trader Joe's and dropping off your kid at an ABA clinic and you've got that three minute, you know, mental breakdown in your car and then you've got two minutes of clarity right after that mental breakdown and you're Googling your kids symptoms, you should land on a website where you can provide very minimal information on your child --  like some symptomatic descriptions and a facial photo because often that's indicative of a specific disease, and some other information -- and then we can run algorithms on that data in real time as you're on the website. If the algorithms think that this person might have a specific genetic disease, then we have a whole telemedicine system built in in the background where we process that information and we manage to get you a test kit within about forty-eight hours of you being on the website. 

You don't pay anything for the whole thing. The test kit goes to you, you collect the samples -- cheek swab or a saliva sample and we're soon adding other biochemical tests as well -- you return that sample to the lab, the lab runs the entire test, and you get a clinical lab report back. The results are explained to you in a telemedicine genetic counseling session. The whole thing is free for patients. We don't collect insurance information, we don't collect payment information. So, that's how the process works for patients. 

The idea is to make it super simple. They don't have to pay and it's really fast. The tests in the background are not actually our own tests. We partner with different clinical labs that can cover different types of tests and we have a pretty large gamut already in the portfolio. Now we can run all kinds of different panels, we can run very specialized genetic tests, we can run whole exome and we can run whole genomes, and then we're working on non-genetic biochemical tests for other types of assays.

Hillary: Wow, and it sounds like you all are just barely scratching the surface here so there's a lot a lot more to come. So, a parent or a family member can go in and input symptoms of maybe their child or family member, they can, you know, upload photos and other information, you run the tests in the background. I'm guessing you're processing tons of data and trying to figure out what's the match, right? And then you're also sending them physical tests that they can do that go to a lab that you partner with. 

So, it’s a lot of different steps in this process. And it is, you know, self reported, and it sounds like there could be maybe potential for error, so talk about maybe the accuracy rates of these tests and what you're finding.

Dr. Lange: Yeah, so I think one thing that's important to explain to people who are not familiar with genetic tests, is that genetic tests are really different from your standard blood iron level tests, or anything that you might be used to. For instance, with an iron level test, there's a number that pops out at the end which is your iron levels and then based on that, a doctor can look at that number and say your levels are too high or too low and based on that we need to do some sort of intervention, like give you supplements or whatever else it is, or you need to change your diet. 

Genetic tests are really different from that, because it's much earlier in the scientific evolution of these tests. When you get a result, when we get the raw data back, then usually two bioinformaticians have to look at the raw genetic data of a patient and sift through hundreds of thousands or millions of mutations -- we call them variants -- and decide which ones of these are pathogenic, likely pathogenic, of uncertain significance, or likely benign or benign. There is an interpretation step involved in this. So already, you've got a level of complexity that's different from standard medical tests. 

The way the system works, we look at what percentage of the people we test get back a confirmatory test result -- so they've got a pathogenic or likely pathogenic variant for a disease that really looks a lot like their phenotype -- and the run lots of different disease programs. For example, in our pediatric seizure program, well over 60% of the people we test actually have a genetic epilepsy. It's a really, really high hit rate and I think if you remind yourself of the fact that these are mostly concerned parents who landed on a website they've never heard of before, typed in some symptomatic information on their kid, and then got a test kit in the mail… that's a really shockingly high number, I think.

Hillary: Yeah, it sounds like it. We know how invested and how passionate and driven obviously, these parents and family members and patients themselves are to really get to the root of their problems. As you mentioned earlier, they've been undiagnosed for seven or eight years, they've gone to a number of specialists, they've probably been on a number of drugs and they're still looking for something that will actually help them in the long term, so that's great to hear. Tell us a little bit about what happens when they do get that positive test result. Okay, we've identified this condition then what happens with the back end and Probably Genetic?

Dr. Lange: Yeah, so the answer is it really depends on the disease that the patients get diagnosed with. In the best of cases, they get diagnosed with a condition for which a treatment is already approved, in which case, the genetic counselors can educate them as to what the approved treatment options are, and what types of specialists can help them get access to those treatments. That's the best case and unfortunately, we only have treatments for about 5% of all rare diseases, and there's something like 10,000 rare diseases out there. So, it's a really depressingly low number. 

Now, the thing that gets me excited is there's a huge research community out there that's conducting treatment development for a very large number of diseases, and that number is only going to go up over time. And because of innovations we have in gene therapies, oligonucleotides, and other modalities, hopefully, really soon, the 5% number is gonna be a lot higher. So, that's if people have a disease that is actually treatable today. 

The second best scenario are people who get diagnosed with a condition for which clinical trials are available, in which case, the genetic counselors in our own system can match those patients to a clinical trial, and we provide them with information as to how they can participate in the clinical trial. 

Then there's, unfortunately, still a really large category of people who get diagnosed with those conditions where we cannot tell them that there is a treatment and we also cannot tell them that there is a trial available. That may sound depressing, and I think it is, but I think it's really important to put yourself in the shoes of a patient or a parent of a patient who spent seven or eight years only knowing that they're extremely sick, and that they're probably getting worse, and not knowing why, right? Those people incur tremendous medical costs, they get treatments that they shouldn't be getting, sometimes those treatments include unnecessary surgeries. For them, even just knowing what the actual cause of all of their symptoms is, is actually a really, really large benefit. And also, once you know what's going on, you can stay up to date on research and development that's actually happening for your disease. So, even if today there might not be a treatment for this, that the answer might be completely different just in a couple of years. 

People might have seen this just a few days ago: the first few kids got dosed with a gene therapy for a congenital hearing loss disorder and that’s a condition where just a few weeks ago, literally nobody had been treated for that. So, I think the good news is there's a lot happening, but the bad news is we've still got a really far way to go.

Hillary: Interesting. We've seen too that when these patient families or patients themselves rally together, that sometimes they're able to identify a researcher who might have done some work on this in the past, but is willing to pick it up again, or they may be able to group together funding and start running their own studies. And so I think it's really a first step in that direction, right? After so much time without information it’s reassuring to know that there's an answer somewhere.

Dr. Lange: Yeah, I think it's actually really important to highlight those patient advocacy groups that you just mentioned. I think this is a concept that's completely unfamiliar to most people, certainly most people outside of healthcare, but even most people outside of rare diseases, where you get all of these crazy parents -- it's usually parents -- sometimes it's patients who go through the diagnostic odyssey, so they've done the eight or whatever years of not knowing what's going on, then they get this terrible message of your kid has this disease and that's it, there's nothing we can do for them, you can just go and make them comfortable. 

A small number of those parents say I'm not happy with that outcome. I'm going to go and I'm going to build a patient advocacy group. I do not have a scientific or medical background, but I don't care. I'm going to go and I'm going to raise money, I'm going to build up this 501c charity and I'm going to try to figure out how the hell gene therapies work. I'm going to try to partner with pharma companies and do this stuff. It's just remarkable how many real heroes like that exist and do that work. We ourselves are partnered with over forty-five different patient advocacy groups who all fit that profile. Many of the successful therapeutic programs we have today literally would not exist if it wasn't for the already grassroots work done by these patient advocacy groups. So, I think you make a really important point there.

Hillary: Yeah, thanks for highlighting that. And I'll just make a short plug for the Year of the Zebra newsletter. Each week we send these out. If you go to osmosis.org/zebra you can get our newsletter that will actually highlight a number of these patient advocacy groups and point you in the direction of donation or support or recent research. So for healthcare students or you know, current providers, or parents or patients, they can get a lot of information there which is great. And speaking of kind of advancing this field of knowledge and trying to improve not just education but obviously therapy and treatment and outcomes for patients…talk a little bit about how Probably Genetic is really improving the rare disease space and where you kind of hope this goes in the next few years?

Dr. Lange: Sure, love that question. I’ve got a short term and long term answer on this. I think the short term answer is pretty obvious, which is, you know, the vast majority of these people go undiagnosed. Being undiagnosed is a really horrible outcome, because not only are you getting treated with stuff you shouldn't be treated with and you don't know what's going on, but more importantly, even if there aren't treatments, or there aren't clinical trials, if you're undiagnosed, you're never going to know about them. So, the short term impact that we're already having today is just to get more people diagnosed, and to get them from all these obscure places in the healthcare system where they're not getting help and get them help. That's the most immediate one. 

The long term vision I think, is even more important, which is, you know, I mentioned that 5% of rare diseases have treatments today and 95% don't have treatments and our hope is that by doing the work we're doing today, and building a really great mechanism to just find these people really, really quickly and get them diagnosed, we can build a platform that in the future, you can do a huge chunk of clinical trial recruitment, treatment, matching real world evidence tracking, and even perhaps most importantly, early stage R&D on because the challenge here is if we don't have really large data infrastructures, where we have DNA symptoms, and contact information of these patients, we just can't crack the nut on a whole lot of these diseases. 

A lot of that was really motivation that came out of the 100,000 Genomes Project, because I think what they're doing for initially a 100,000 people, and now a few million people, I think if you spun that off to a global level and built a really large platform, you could really usher in a new age of therapeutics we probably just can't get to right now.

Hillary: That's a very exciting future, and hopefully something that accelerates in the coming weeks and months. Interestingly, I think with COVID, we've heard from a number of researchers that that kind of shifted the way that research is done on a global scale and how information is shared and hopefully some of the infrastructure that was built during that time could be leveraged for other things such as rare disease research and therapeutic advancements. So, we'll certainly be following that. 

You mentioned one story of Julia the parent that you met earlier, but are there other kind of lasting imprints of patients that have been impacted by Probably Genetic or maybe stories that you can share about its impact?

Dr. Lange: Yeah. My fiancée always tells me that I'm not the best party guest because I've only got really depressing stories, but I tell them regardless. I'll just pick a few. I mentioned the epilepsy program. You would think that we would only diagnose patients who don't have access at all to specialized health care providers who just know what a genetic epilepsy looks like. But actually, many of the patients we diagnose are being seen by chairs of pediatric epilepsy at leading universities in the United States. We've got patients who are being seen by board certified neurologists, we've got patients who are being seen by members of the American Society for Human Genetics. Unfortunately, they were actually seeing the right specialists and still didn't get diagnosed. I guess the way to put a good spin on that is we found them and now they got diagnosed. So, I think that's a plus.

I'll give you two really depressing stories just to kind of round this off. We have one patient that we diagnosed two weeks after they had a leg amputated because of their genetic disease that had gone undiagnosed up until that point. That really shouldn't happen like that. We just have to be faster with that and we have to scale it up a lot so that we can find these people a lot earlier. And mind you, that person had an amputation without knowing why they're having an amputation, and that's just completely crazy.

Another problem that we're working on right now is a condition called osteogenesis imperfecta, where patients have many bone fractures often really early in life. And one of the really depressing things about that is the most common misdiagnosis for children with that condition is child abuse. So, not only are their parents going through this experience of ‘my kid is breaking their bones all the time, I don't know what's going on, and nobody's helping me’ but all of a sudden, they have to fight the system to not lose their kids because people are suspecting that they're hurting their own children. That's the depressing side of it. But the hopeful side of it is if you find those patients really quickly, you can get them diagnosed and hopefully, you know, avoid worse outcomes. 

Hillary: What a stressful situation. These are very sobering stories, but obviously we hear the other side too and it sounds like as you continue your work we’ll hear many more positive stories, hopefully, of people finding their diagnosis sooner and getting the right treatment sooner. 

I know we're running short on time, so I just have two quick questions for you. Osmosis is an education company and we love to fill in gaps in knowledge. What areas would you like to see Osmosis create more educational content around?

Dr. Lange: I think one that you could do -- and I don't know if you've done that, so apologies if you have -- is I think there needs to be a redefinition of what ‘rare’ means. Because yes, every single one of these ten thousand disorders is rare, but collectively, that's something like four hundred million patients on the planet. So, given that a whole lot of medical students follow your content and read everything you put out, I think medical students should really switch their mindset and instead of thinking, this is the odd one out, you know, that's the zebra, they should consider that it's extremely likely that they'll encounter many really unique cases in their journey as physicians. So, that's part one kind of reframing that and changing their mindset. The rare thing is probably the most likely thing you'll encounter. That's part one. 

And then part two is, they'll never stand a chance to be able to diagnose those 10,000 conditions on their own, because they don't have enough time in medical school with all the follow on workshops and everything to learn the diagnosis of those conditions. But if you can become really smart about using tools -- like the stuff that we're putting out and like the stuff that other artificial intelligence companies are putting out -- then you can get assistance from technology to solve that diagnostic challenge that you yourself simply can't possibly master because it's just it's too much information. So, those two: reframing and technology. 

Hillary: I'm guessing that yeah, students or healthcare providers can access Probably Genetic and could they potentially go in and input information on the patient's behalf and get that information? Is that how some are using it today? 

Dr. Lange: Actually, yeah. We didn't intend like that originally, but we can already see that tons of physicians are using it that way and we're working on building more information flow so that the physicians can stay up to date with the process that their patients are going through as well. 

Hillary: That's great. Well, hopefully we see more of that. I know I'm leaving you one minute here, but maybe just for those listening, do you have any advice for future clinicians or existing clinicians as they encounter some of the challenges in healthcare today?

Dr. Lange: That's a tough one. I mean, from where I'm sitting, I would say familiarize yourself with everything that's going on in clinical development for these conditions because I think the conventional wisdom that if there is no treatment today, it doesn't make sense to try to help the patient get diagnosed is not only wrong, I think it's harmful. So, stay up to date with what's going on in research, even as you venture out and leave academia because there's a lot happening and the stuff you learned today is not gonna be treated more than five years from now.

Hillary: That's great advice. So, Dr. Lange, thanks so much for being with us today.

Dr. Lange: Thank you so much. This was fun.

Hillary: I'm Hillary Acer. Thanks for checking out today's show. Remember to do your part to raise the line and strengthen the healthcare system. We're all in this together.