Improving the way to improve cancer using AI
Episode 163
Thanks for tuning in to the Outcomes Rocket podcast where we chat with today’s most successful and inspiring health leaders. I want to personally invite you to our first inaugural Healthcare Thinkathon. It’s a conference that the Outcomes Rocket and the IU Center for Health Innovation and Implementation Sciences has teamed up on. We’re going to put together silo crushing practices just like we do here on the podcast except it’s going to be live with inspiring keynotes and panelists. To set the tone, we’re conducting a meeting where you can be part of drafting the blueprint for the future of healthcare. That’s right. You could be a founding member of this group of talented industry and practitioner leaders. Join me and 200 other inspiring health leaders for the first Inaugural Healthcare Thinkathon. It’s an event that you’re not going to want to miss. And since there’s only 200 tickets available you’re going to want to act soon. So how do you learn more? Just go to outcomesrocket.health/conference. For more details on how to attend that’s outcomesrocket.health/conference and you’ll be able to get all the info that you need on this amazing health care thinkathon. That’s outcomesrocket.health/conference.
: Welcome back once again to the outcomes rocket podcasts where we chat with today’s most successful and inspiring health leaders. I welcome you to go to outcomesrocket.health/reviews where you could leave a rating and review for our outstanding guests today. His name is Peter McAffery. He’s a Biodesign Fellow at the Texas Medical Center Innovation Institute and a TMC Innovation Institute. And he’s doing some really great things. He’s a physician a data scientist a medical inform assistant entrepreneur who believes that data fluency and artificial intelligence have the power to heal healthcare and the lives of patients it serves. He’s doing some amazing things his background as a as a physician and clinical pathology from the Harvard Medical School and he’s just got some really passionate beliefs and is doing what he believes through his passion in this program. And I want to introduce this amazing gentleman to the podcast. And so Peter I wanted to welcome you.
: Thank you, Saul. It’s great to be here.
: Absolutely. So so maybe you want to fill in any of the gaps here that I may have introduced in intro.
: Sure I know you a great job I was interesting. Trivial fact. I come from a humanities background. I don’t have formal training. Science we can get into it or I think that gives a different kind of a unique perspective when you approach those kinds of skills a little bit later in life and how you use them to find value. But I’ve been involved in both operations, analytics and healthcare analytics as well as bioinformatics I’ve kind of done a couple of different things and had had three companies in the past have been involved with. So a lot of interesting war stories to draw from
: Love it. And they were excited to go through that. Never a dull moment with Peter McCaffrey. Ladies and gentlemen and so Peter that’s pretty cool them humanity side. I also well my my undergrad is in classical humanities. What was yours?
: Wow and I was an English literature and philosophy.
: I love it man. And everybody is always like why. Like why. How are you in healthcare. And so that lends very well to our next question. Peter, what got you into the medical sector?
: That’s a great question so I started out deliberately not wanting to be in health care and I was in high school I was I wanted to be a musician. I was more a creative type and I was always told this’ll probably resonate with you and perhaps some of your listeners that you know I was always told that there is this split you know that people who are creative do arts and people who are quantitative science and they’re kind of just separate things. I can testify that’s not so true. You know I got into humanities because I just I liked it and it seemed open ended. But what it really teaches you how to do I think is to think about things of philosophy as well literature you know your skill isn’t analyzing stuff. And so it was with that joy of doing that loving that kind of thing that I had. I decided to try a couple of health care experiences you know shadowing I went on a couple of medical mission trips and it really struck me that healthcare when approached the right way is a really analytical field. You can be creative. It’s not just number crunching. So seeing I could add my interest in analysis with impacting real people’s lives. I thought this is great and I haven’t looked back since I was a good decision to go into.
: That’s awesome Peter and I love the way that you kind of laid that out. You don’t necessarily have to. If you’re artistic you don’t have to stick with the art you could go into science and you’re a great example of that and I never thought about it that way you know we spent a lot of time analyzing Greek literature and like you had to in order to make sense of it you have to be analytical.
: Yeah absolutely. And to formulate arguments you’ve got to put evidence together and make conjectures you have to articulate them just like he should do many things.
: That’s so cool. So really glad and listeners I know you remember we had Joowon Kim from the TMC Innovation Institute. She and Peter as well as a couple of other folks are working on their company there and so Peter, thank you so much for making the time to be on here. Oh absolutely. Thanks for having me. You know yeah it’s fun. So tell us a little bit about what you guys are working on. Peter I know the episode with Joowon we launched it. She was talking about kind of how you guys narrowed into an idea so maybe you could take us a little bit more into what you guys have evolved into.
: Absolutely. So the two second recap of kind of how we got here is that the as bad is on fellows. We spend the year July – July identifying unmet needs prioritizing down and then finding one and then start trying to solve that basically. And the needs that kind of bubbled up to the top for us both because they’re important but because we liked them and thought they were interesting and a mixture of a lot of things were really precision medicine oriented needs. And so we kind of focus on the problem of being able to identify patients who will respond better to cancer therapies kind of as our as our first problem. But we solve it in a way that we may have the ability in the future to help patients respond better. So it’s predictive. We can also do therapeutic things and we do this through a focus on basically DNA analysis using machine learning.
: Super interesting and this is a hot topic. A lot of people are talking about precision medicine and you had several guests talking about it. What are you envisioning Peter will be this is a hot topic right this is what you’re thinking is the hot topic so how are you guys addressing it differently with your company. And what are the outcomes a potentially get from what you’re doing?
: That’s a fantastic question. So traditionally and this is near and dear to me as a as a pathologist you know traditionally we thought of cancer as these big buckets that were kind of largely defined by what things look like or where they came from. You have a lung cancer, you have a skin cancer, you have kidney cancer. And we’re realizing that’s not really accurate your cancer is very personal. It’s very unique. It’s really described by a whole lot of measurements about genomic and what they look like and what your history is. So our kind of unique angle on that is that recent research has shown kind of yet another angle through which cancer can be looked at and through which its behavior can be described. And that’s through what’s called the gut microbiomes I can get into that in a second there. But we look at these bacteria that live in patient’s guts that they basically do things to the patient’s immune system do things the drugs that they take. As we look at the DNA of those bugs and we say what’s going to happen to this patient with their drugs. Can we change that microbiome? Can we make them do better? Can we achieve better efficacy for more people?
: So you guys are looking at. And that’s fascinating right. I feel like there’s been a surge of focus in the microbiome and now you guys are taking a pretty cool shift here in precision medicine for oncology. What specifically are you guys doing? Can you get into some of the specifics on how you could improve outcomes there?
: Absolutely. So it’s kind of an interesting story in the way we look at this problem cancer care has changed a lot in a lot of ways. Used to be scalpels and radiation and generally kind of poisonous chemos and that still exists in large part. Then you had all the targeted therapies people cannot with their antibodies against certain things and now kind of the new I guess pillar or the new domain of cancer care is largely what they call immuno oncology. And what’s interesting immuno oncology is you. You’re trying to use the patient’s own immune system to actually attack cancer because you’re facilitating the immune system to go and do the work of finding the cells and beating them up and suppressing the tumor as it goes forward. So it’s this kind of new wave of use in the immune system as a weapon against cancer. The microbiome it turns out plays a big role in how your immune system will behave. So what these bugs are what they are what they do how they act how they interact with your gut teaches your immune system basically to behave a certain way makes it active, makes it inactive. So we basically are an immune system profiling tool that allows you to make the patient more responsive to these immunotherapies.
: Fascinating fascinating and I’m assuming you’re taking samples from the gut like how are you guys taking the samples.
: Yeah so we use fecal samples so you know they get micro biome is one of the very easy things to sample about a patient you take stool which has billions of microbial organisms in it across thousands of different species and you basically isolate all the DNA from that stool and that is just like a soup of DNA from all these organisms. The whole system in one sample and you sequence it and you end up with this just massive data dump of fragments of DNA from thousands of species of microorganisms that are all doing things and an ecosystem and you try to figure out from that who is there what are they doing. How do they relate to one another and what are these functions that they might do in this patient is driving their disease or driving their response. We rely heavily on machine learning to do that particularly in neural network and artificial intelligence analysis of those data to kind of get those patterns. But it’s really pattern recognition of this very kind of broad shotgun sample of data from stool.
: That’s fascinating Peter and and it sounds promising Yeah I was talking to a friend of mine and we got into this pretty deep conversation critical care. And how samples of like sublingual CO2 are just as effective at understanding patient’s condition as if you were to put in I forget the name of that tool but you put it all the way into the stomach and it’s just like all tied together and it’s super intriguing that you guys are piecing this together to find out the efficacy of these drugs.
: Yeah it’s fascinating because you know we’re realizing that the microbiome is like it’s like a whole new organ. If you look at you know a human being if there’s take a person your genome has like 23 24 23000 genes basically. And yet your microbes that live in you and only you have over 3 million. So if you as a sort of whole organism are mostly bacterial genes and yet we don’t know much about what these do and we don’t know much about how we can like take control of their behavior and to achieve outcome. And it’s ironic because you know Hippocrates thousands of years it goes you know all. What is something like all disease begins in the gut or something like that. And it turns out that unbeknownst to him exactly how that was true. That may in fact be much truer than we thought.
: That’s pretty cool man. And that’s kind of like how the humanity is here intersects with science and your perspective is really cool and I love this approach that you guys have taken just using new technologies and applying them in a way that will help treat these these very dangerous diseases. Cancer is a big deal. You know so many people die of it and if we have a more effective way of treating it you’re making the world a better place. So kudos to you and your team my friend.
: Appreciate that. Thank you.
: Absolutely. And so we’ll open up here for some more a little bit later listeners I mean definitely want to dig into this company and the things that they’re up to. But I wanted to dive into some of your stories. You’ve been through several different companies you’re doing a lot of great things in medicine. Share with the listeners a time when you failed and what you took from that lesson.
: Absolutely. So a harking back to kind of to other companies I worked on one was a a life sciences company was 2011 where I was the sole founder and CEO of a company called Accetia and we made a tool. We don’t operate anymore with a tool called omic where you kind of design analysis pipelines for DNA in the cloud and run these analyses and then the analysis parcels in the cloud and it was all very exciting stuff. But I remember very distinctly managing that process to really important lessons came to mind. No we got funding to build a certain spec of this tool. We had deadlines, milestones, burn rate all those things although it ended up being okay when we finally were able to produce it and render it on into the institute we were partnered with. I did not do a very good job in my own opinion as identifying what of the many features was most valuable. And I think for anyone who builds stuff or manages people this is a hard skill to get right and a very easy skill to get wrong. I spent a lot of time saying well we can do all these things you can have all these features we can have you know this widget does that and this widget does that did not do a good time a good job of sequencing when you need to nail those things. And what is really important is you have to pare down things you do to be what are. If I could only pick two or three things that worked, what are those things going to be? Let me do those now only once those are done. You think about the other ones. And let me continually validate whether the things are valuable just because I think they’re cool doesn’t mean that they’re worth anything. I think that’s a I’ve gotten that on a number of times. That is a very important lesson to learn as you got to be open to externally validating get people’s opinions. Pace yourselves and focus on sure fire easy wins upfront. Put fluff on the back burner.
: That’s awesome. And what a sustained way of saying it Peter. Definitely a good take away listeners. We all are guilty of falling in love with our ideas. And in this area that we all work in of health it’s so important that we huddy to Peter’s point because if we don’t a lot of time flies by your burn rate just will put you out of business eventually. But Peter your idea of just focusing on that one or two things if I had to throw everything else out the window. What one or two things will I focus on?
: Yeah that’s key. It’s tactical.
: And so these invaluable experiences. Peter I’m sure you’re bringing into your new role here with the new company. Absolutely. What would you say some of the greatest things you’ve done and you’re most proud of in medicine to date?
: So this would be kind of seems to be two years ago. Yeah so when I was a resident. A co-resident one of my good friends John and I we started a company called Hadera technologies. And our goal really was not bioinformatics it was operational analytics so we were you know clinical lab people we sit in the core laboratories of the hospital so it’s a nice vantage point from which to do things like ops analytics because you see all the tesco and you see them all go out you see turnaround time you see ordering behavior right. So we thought this is really good let’s build a dashboard, let’s build a platform where you can easily get metrics and turn around and look at things and predict things. We didn’t have a great insight into how complex deploying operational software in a healthcare setting really is and it was an amazing journey to say from when we decided we want to build this to fight it out a year and a half later. What I was really proud of us for is really going through getting stakeholders interested involved building will not cutting corners not going behind the back of a process doing it the right way through you know your security validation through your beta and pilot getting your spec looked dead by all the health I.T. department all the things that you have to do to really do healthcare I.T. and roll out something in a hospital. We kind of spearheaded and did ourselves and it was painful but it was great to do it and see that thing actually get turned on and it’s still running and known and still does its job there and I was proud of us for pulling it off because you know deploying things in a hospital is not just building, building’s critical but it’s political. You have to be able to work with the organization you have to able to work people.
: That’s for sure and layering in all the hipa and policies and safety you know cyber security is a very real thing. This is great. And kudos to you and your and your friend. Now you guys it’s your legacy there right which you’ve done will continue to help them even though you’re not there anymore.
: Right exactly.
: That’s pretty cool. And listeners. This is another point right. If you look Peter was in the facility. You know he’s a physician in the facility and it still took him and his friend a year and a half to get this done. Now I hope that tells you something. I believe that we could get a lot of unique thinking from outside of health care and we need it right we need it to make it better. And at the same time, don’t think that you could build a business and be successful without the help of clinicians or the people your stakeholders. That’s the number one mistake. Peter obviously you did it and it still took you all.
: Absolutely yeah. You have to work through people in hospitals as you know they are huge organizations they have a lot of moving parts and engineering is only one it’s important. I would never devalue it but it’s only one piece of very large pie and you got to kind of see it from the perspective of some other people too. I myself had a tendency certainly to be frustrated saying well you know what’s the problem I built it it runs past tests. Why can’t we just turn on. You don’t consider that like from the perspective of the I.T. security guy this is the eight hundredth app he has to roll out and he didn’t build it. It’s not his baby. You got to get him invested enough to help him move forward. You got to work with people you know.
: Absolutely. And if you’re a tech leader listening to this episode I am sure that Peter would love to connect with you and maybe share some ideas to help your company impact health in a good way.
: Absolutely.
: And then at the end of the podcast Piero’s share his contact information so that you have it. Tell us about an exciting project. Let’s go back to the idea that you guys are working on this business of the biome ans oncology. Give us an exciting focus that you’re working on today with.
: Absolutely. So we’re looking at how we can use the microbiome itself that profiles and signatures in the microbiome to tell whether an individual patient will respond to immune checkpoint therapy which is a category of chemo. And it’s really fascinating because the way we were. What I like about it kind of geek out on the moment is we are we we are using neural networks to do this and I feel like machine learning and AI are very hyped and there’s a lot of misapplication of that. But what’s really fun about this project is that there’s a genuine use case for why or why it would apply here. And it’s just really exciting to even in what we’ve done so far. Prototyping to run these analyses watch that they do learn patterns and watch the patterns are externally validated in the literature. It’s it’s amazing that it mixes engineering and science and that kind of infectious like oh my goodness it actually worked aha moment you have when you when you first build something you see it complete. That we we believe the business case is good because cancer is a real and pressing illness and that these checkpoint drugs offer a lot of value and need to be enhanced. But we all just love. We love the tech too. We love working on it and we love watching the code deploy and watching the analyses run. It’s just fun to do.
: Yeah. And even more fun part Peter will be when you guys start applying it and start saving lives.
: Absolutely. Yes. That brings it all full circle.
: That’s amazing man. Thanks for sharing that.
: Of course. Thank you.
: So Peter getting to the end here. Let’s pretend we’re building our medical leadership course on what it takes to be successful. The 101 of Dr. Peter McAffery. And so we’ve got a syllabus here for questions and at the end I’d love to hear about a book and a podcast that you recommend to the listeners. You ready.
: Absolutely yeah.
: Awesome. What’s the best way to improve health outcomes?
: I think you have to measure that to measure will measure everything you know where you are.
: What’s the biggest mistake or pitfall to avoid?
: Honestly I think it’s a tendency to have this reliance on blackbox tools. You can’t Hirth analysis to just a tool you buy it and leave it at that you have to understand what’s happening.
: I love that. How do you stay relevant as an organization. Despite constant change?
: I think you have to focus back on fundamentals. I think you have to make especially as physicians for medicine. You got to commit to knowing sharpening your skills knowing your chops knowing analysis knowing how to look at measurements and make conclusions and you stick to that. Everything else is subservient to that you’ll be good.
: What’s one area of focus that should drive everything in your health organization?
: So the way that we like to say it for us is you have to question everything and try anything. Never assume because a general opinion says something what works. But it won’t work. You’d be surprised.
: That’s beautiful. What book and what podcast would you recommend?
: Yeah so for podcast Cloudera. If you’re into big data data science Cloudera has a really really great podcast we talk about cool their tech people from industry for a book it’s interesting. One of my favorite books ever that I read in medical school actually is Dune Frank Herbert’s Dune sci fi novel.
: Dune?
: Yes. Okay. The first one just because it tells a story of a guy who has all these what look like magical powers and it turns out that it’s not magic in that universe. The guy has an ability to take information synthesize it and see the future. I think that’s what we should think about health care like there’s no magic here. There’s skill and there is knowledge and we can all get there we can all do that.
: What a great great thought there Peter and listeners don’t worry about writing any of this down. You could get links to Dune, Cloudera as well as all of the Q and A that we’ve just been chatting here with Dr. McAffery. You can find all of that at outcomesrocket.health/PeterM.. So Peter take us down here. Bring us to a good close. What would you say to the listener as a closing thought and then share where you could best be contacted.
: Then my closing thought would be medicine for all the great things that have been done in medicine and there’s so much more that hasn’t been done. And today could be the beginning of a life changing project for anyone listening to this. Go out there Google what you’re passionate about. Learn a bit about it. Roll up your sleeves and get cracking and you can get farther than you ever thought. I can testify that.
: That’s awesome.
: For my e-mail address. I guess my main would be pmccaffrey@tmc.edu
: Outstanding and will also include links here to Peter’s link then anything on Twitter or anything else like that that you like for listeners to know about.
: I should have a twitter. I don’t.
: I actually didn’t have one. And as I was building the podcast Peter, I had a couple of friends notably Nick Adkins from the pink socks movement he’s like oh you just need a Twitter account period.
: Yeah.
: That’s a good share space.
: I have to make one.
: Yeah. No worries. So listeners Well again we’ll include Peter’s email and his linkedin. Feel free to connect with him. Just go to outcomesrocket.health/peterm and you’re going to find all of that as well. Peter, this has been such a pleasure. Thank you so much for sharing your words of wisdom. And we’re really excited to see how the company turns out.
: Wonderful. Thank you for having me. This has been wonderful. I appreciate it.
Thanks for tuning into the outcomes rocket podcast if you want the show notes, inspiration, transcripts and everything that we talked about on this episode. Just go to outcomesrocket.health. And again don’t forget to check out the amazing Healthcare Thinkathon where we could get together took form the blueprint for the future of healthcare. You can find more information on that and how to get involved in our theme which is implementation is innovation. Just go to outcomesrocket.health/conference that’s outcomesrocket.health/conference be one of the 200 that will participate. Looking forward to seeing you there.
Recommended Book and Podcast:
Best Way to Contact Peter:
pmccaffrey@tmc.edu
Episode Sponsor:
