AI in Veterinary Medicine | Dr. Kelly Diehl

[00:00:00] >> Molly Jacobson: Artificial intelligence is already changing the way veterinary medicine is practiced, and this can benefit your dog today.

[00:00:09] >> Announcer: Welcome to Dog Cancer Answers, where we help you help your dog with cancer.

[00:00:15] >> Molly Jacobson: Thanks for joining us, listener. I’m Molly Jacobson, Editor-in-Chief of DogCancer.com. Artificial intelligence is transforming our world and veterinary medicine.

[00:00:28] Today, our host, James Jacobson, founder of the Dog Podcast Network, is joined by Dr. Kelly Diehl of Morris Animal Foundation. Morris Animal Foundation has led the pack in funding veterinary research since since 2008, and their Golden Retriever Lifetime Study has proven invaluable for our understanding of dog cancer.

[00:00:51] In their conversation today, Kelly and James discuss the Foundation’s newest findings. And as they get into it, a fascinating trend emerges. AI is not just a tool for humans, it can help dogs, too, by detecting subtle patterns in their health data that helps catch disease early. In this wide ranging discussion, you’ll learn about how human cancer researchers are integrating their work with veterinary cancer researchers.

[00:01:19] How AI is bolstering the one medicine approach to healthcare. The potential for AI to quickly uncover new insights into cancer and other diseases. Speculation on how AI might impact veterinary practices, both in the long term but also the short term, especially in early diagnosis. And, of course, the challenges that come up when you can now detect diseases earlier than you’ve ever been able to detect them before, and you don’t have protocols for how to treat at those early stages.

[00:01:54] Today’s show is for you if you’re a dog lover, an AI enthusiast, or if you subscribe to the idea of One Health, which is looking at how animals, people, and the environment all impact each other’s health. Now let’s go to James Jacobson’s conversation with Dr. Kelly Diehl of the Morris Animal Foundation.

[00:02:17] >> James Jacobson: Dr. Kelly Diehl, thank you for being with us today.

[00:02:20] >> Dr. Kelly Diehl: It’s great being here. Thanks for having me on.

[00:02:23] >> James Jacobson: It’s great to have you here today. We have been covering the Morris Animal Foundation since 2008, I think. It’s been a while and uh, it’s always good to get an update on the amazing work that you guys are doing and it’s just expanded beyond the lifetime, you know, Golden Retriever Study to so many different things, but really the focus continues to remain on cancer, right?

[00:02:48] >> Dr. Kelly Diehl: Yeah, we’ve been a leader in cancer funding for a really long time in a variety of different species. You mentioned the Golden Retriever Lifetime Study, which is probably our deepest dive into cancer.

[00:03:01] >> James Jacobson: Right.

[00:03:02] >> Dr. Kelly Diehl: Obviously, in dogs. But we’ve been doing it for quite a while, but yes, we do a lot of cancer research.

[00:03:10] >> James Jacobson: And you have been expanding. I saw this year you have eight new initiatives that you have funded or in the funding stages for.

[00:03:18] >> Dr. Kelly Diehl: Right. We have a bunch of new grants. We, a couple things, one is we did a call for hemangiosarcoma specifically, which is, I’m sure your listeners know, a really bad cancer of dogs. It affects a lot of dogs. It’s a horrible disease. It has a terrible prognosis. And so we decided to focus on that and focus fundraising around that. So we have several studies that fall under the quote, hemangiosarcoma initiative bucket. But we also just went through our canine call for proposals.

[00:03:51] And this year, again, we focused on just canine cancer and not other topics, which, you know, we fund lots of other things. So we have, even more studies now going in the canine cancer bucket, so quite a bit. And of course, the Golden Retriever Lifetime Study at its heart is a cancer risk factor study and that’s been going on since 2012.

[00:04:16] >> James Jacobson: Right, we’ll talk about that because that’s the legacy. That’s the thing that I think most people are familiar with and it’s kind of cool because you’ve been looking at these dogs throughout their lives. But some of the stuff that you know caught my attention are things like artificial intelligence, and how that could actually help with dog cancer. Let me start there. What are you doing there?

[00:04:38] >> Dr. Kelly Diehl: Yeah, this is a new area for us, as it is for a lot of people, but I think we all realize that AI can pick up patterns maybe we can’t as humans. Very subtle patterns in blood work or test results or even x-rays that we just can’t see. Uh, we’re funding two studies that are using AI in different ways.

[00:05:01] One study is going to try to leverage AI to look at lymphoma. Are there factors, something in blood work that we’re not seeing that might point toward this dog, maybe at risk for lymphoma, which, you know, we talk a lot about epimangiosarcoma, but lymphoma is the most common, you know, cancer in dogs and cats too.

[00:05:25] >> James Jacobson: Right.

[00:05:25] >> Dr. Kelly Diehl: So that’s very exciting. The other one is interesting because I think we all know, or have heard, about AI just having better eyes than we do, right? They’re using AI to screen radio x-rays in ways that maybe people just can’t see stuff. And we have a researcher that’s going to do that with slides. So we, you know, typically with lymphoma, we aspirate a big lymph node or a spleen or something, we squirt it out on a slide. And in general, we have a really good chance of detecting that.

[00:05:55] >> James Jacobson: But it relies on a human being to look at that and to interpret the.

[00:05:58] >> Dr. Kelly Diehl: Right, right. And sometimes it’s a no brainer, and sometimes, though, it can be more subtle. And we also know there are subtypes of lymphoma. So we have someone who’s going to try to see if looking at slides, they can first train, right? Because AI has to be trained too. It’s just not going to like, you can’t pop a slide in there and say, is it lymphoma or not? It has to be trained.

[00:06:19] And then, can it learn and be trained to pick up maybe subtle changes that we, as humans, looking at these slides could not. So those are some really exciting ways that we’re interested in harnessing AI. And I think one of the things that I’m really excited about is that people are submitting stuff. Right? We’re starting to see it trickle in different ways to harness AI, even for other diseases as well.

[00:06:51] >> James Jacobson: And you’re not seeing it just from veterinarians. I mean, I looked at some of the people who are getting grants and they’re not strictly in the veterinary space.

[00:06:59] >> Dr. Kelly Diehl: Exactly. And we’re really excited about that too. I think spreading our message, I guess you could say. We have people from.

[00:07:07] >> James Jacobson: And money.

[00:07:08] >> Dr. Kelly Diehl: And money. We have people from Harvard. We’ve had applications from Yale, Duke University, and again, human in quotes, air quotes, typically, we think of them as human. They don’t have a vet school necessarily associated with them, but many people are recognizing that animals are good models for human disease because we don’t give them the disease like you might like, here, I’m putting a tumor in a rat. Well, that’s not how tumors go.

[00:07:35] I mean, it’s a useful model, but not, but the benefit too for dogs is they benefit from this interest, right? There’s finances. There are really smart people who are joining our veterinary core, which is really small. Right? Veterinary cancer researchers is a small group. We’re expanding.

[00:07:56] >> James Jacobson: Oncology in general in the veterinary community is, is ridiculously small. And then obviously a subset of that are the, you know, we have the practitioners and then we have the people who are doing the research and that’s even smaller.

[00:08:07] >> Dr. Kelly Diehl: Right.

[00:08:07] >> James Jacobson: So AI is attracting these great minds to look at it because I would imagine from a macro perspective, we get into this concept of one medicine. Like, if it works in a dog, you know, a cancer in a dog is often similar to cancers in humans.

[00:08:23] >> Dr. Kelly Diehl: Right, exactly. And vice versa. I think it happens a lot vice versa too, which is, can we use drugs, right, that have been developed in people, but if we can maybe interdigitate that a little bit closer, rather than, okay, a drug gets developed in people and then it trickles down to veterinary medicine, can we be closer to that discovery and development than waiting a couple years and picking something up, right, and, and doing it?

[00:08:52] And I am really excited that we are seeing that. That what we’re doing is much closer to the human side, and we’re going to benefit, our dogs are going to benefit a lot from that.

[00:09:04] >> James Jacobson: Yeah, that’s pretty cool. Interdigitate. That’s a good word. I like that.

[00:09:07] >> Dr. Kelly Diehl: Yeah.

[00:09:08] >> James Jacobson: I’m intrigued with the one that is basically trying to possibly train AI to be, a radiologist because I mean, so many people who have had dog cancer get the reports back and it’s like, it’s one of those, well, unremarkable, kind of remarkable, little bit, you know, there’s a very large wishy washy area that so many of our listeners and viewers have experienced. And I’m sure as a general practice that it just is confounding, right? So having the potential for AI to be a little bit better is pretty cool.

[00:09:42] >> Dr. Kelly Diehl: Right. Especially with something as inexpensive, typically as putting cells on a slide and sending it in, right? Because we know we have pet insurance, but it’s not like us, right? And I, as a practitioner, boy, I had clients, if they didn’t have some insurance, they would have never come see me as a specialist, right? They just couldn’t have afforded it. But still, you’re going to pay a lot out of pocket.

[00:10:05] So if we could leverage a really simple, cheap thing like aspirating a node, blowing it out on a slide and sending it into a lab, and they can tell us tons of information from that without additional testing, wow, that would be, that would be really great. We had one recently also submitted that is, okay, so we always talk about sometimes x-rays not being very helpful, right? In some cases, but they’re cheap, they’re easy. People have it in their practice.

[00:10:34] And there are people who are now looking at, well, maybe, maybe we should take a look at x-rays, which we blew off as saying, Nah, we can’t see anything. But give it to AI and train it and maybe it can pick up subtle differences again that our eye, even with all the technology, if you’ve ever seen digital x-rays, it’s really fun. You can like zoom in. You can zoom out. You can call. I mean, it’s, it’s been a great advance versus the old days where you’re throwing the slide, the x ray up on the thing.

[00:11:04] But suppose AI can look at that for us and say, you know, you can’t see this, but I can detect the subtle change and you need to pay attention. Are we going to find get information that would otherwise be lost to us? And again, something a general practitioner can do in their clinic. X-rays aren’t super cheap, but they’re cheaper than maybe getting referred for an ultrasound, right?

[00:11:28] >> James Jacobson: Right. And a lot more readily available, because as you say, every vet practice has it.

[00:11:32] >> Dr. Kelly Diehl: Absolutely.

[00:11:33] >> James Jacobson: And I would imagine just to train the AI, there’s an enormous amount of data that is housed on servers somewhere that can be given to the, to train it.

[00:11:42] >> Dr. Kelly Diehl: Yeah, if you don’t mind if I deviate too slightly from cancer is in the UK, they’ve pioneered this because there are many, many of their vet clinics share a similar software. So if we go to a hospital, like a big hospital system, right, like Kaiser, I’m just going to throw that out there. You could be Kaiser anywhere and they’re going to have access to your records, right? We have more organized and similar, partly because of insurance, right? You have coding, most vet practices, not integrated at all, even across veterinary hospitals in the United States.

[00:12:18] But in the UK, they often use the same software. So people in the UK are training AI, they’re looking at a disease called Cushing’s disease to pick up because that can be a real boogaboo to, to find early and the earlier the better. And they’ve had some success training AI to look at very subtle changes in blood work, et cetera. And then the AI contacts the vet and says, Hey, Hey, this dog may have some signs of hyperadrenocortisism, have you thought about it could be an early case?

[00:12:54] And again, it’s because they have this integrated system. They’re looking lots and lots of data, but suppose we could do that for cancer and, uh, drum roll please, we actually have a project with the Royal Veterinary College in London, their vet compass group, and guess what they’re going to do? They’re going to look at some data from the UK. on cancer, and they’ve also got study data from us.

[00:13:20] >> James Jacobson: Are they looking at x-rays? Are they looking at blood work? What are they gonna be looking at?

[00:13:24] >> Dr. Kelly Diehl: Just data in general, like what does, because we have obviously a lot of lifestyle data and blood work results and things like that. So they’re, they’re coming at it. I wouldn’t call it AI per se, but they’re the AI folks. So they’re just trying to like, just get a big giant database. Okay. So we have this database of dogs in the UK. Here we have this database of Golden Retrievers in the United States. And they develop cancer and they’re being followed really closely.

[00:13:52] How about if we look at this big group and see if we can compare and find some commonalities that might actually identify real risk factors, not just in a closely studied group of dogs, but in a more, a bigger group of dogs. So it’s some exciting stuff, but their, their really risk factor, you know, is part of their thing.

[00:14:11] >> James Jacobson: To me, it’s, it’s exciting. I mean, I get chills thinking about it because for the longest time we’ve joked that, you know, DogCancer.com will be here forever, until they find a cure for cancer, but with this really interesting stuff and early detection and the ability to amass all this data that is there, it is possible, you know, sometime to really make giant strides because the computer is going to figure something out. So they’re going to be looking at, In Great Britain and the studies that, are you helping to fund the British study?

[00:14:40] >> Dr. Kelly Diehl: We are a little bit and Boehringer Ingelheim is also.

[00:14:44] >> James Jacobson: A big animal pharmaceutical company, animal health.

[00:14:47] >> Dr. Kelly Diehl: Right, they kind of jumped on board too to help fund it, but a lot of it is coming from us. You know, I think of AI, and I also think of big data, and this is kind of big data. It’s more of a big data project. But it is, you know, again, I see them as kind of, again, interdigitated, right? And big data can really help us if it’s analyzed properly. It can really make a difference when you have thousands and thousands and thousands and thousands of dogs, right that you can look at.

[00:15:17] >> James Jacobson: Right.

[00:15:18] >> Dr. Kelly Diehl: And that’s the first time we’ve been able to really do that.

[00:15:20] >> James Jacobson: And do you see it having an impact in this interdigitation with all the data that you’ve already collected through say the Golden Retriever Study to look at I mean, you have the data, but right now it’s being analyzed ostensibly by human beings, right?

[00:15:35] >> Dr. Kelly Diehl: Right, right.

[00:15:36] >> James Jacobson: And the training models in AI to look at, oh my gosh, they were eating this dog food, and it was this supplement, and it was, I don’t know, they used this lawn chemical company to, you know, take care of their yard, and then we see a higher rate of this or something like that.

[00:15:55] >> Dr. Kelly Diehl: Right, right. So all of that, because I think, what we hear all the time, and I’m sure you hear it, is what’s the best diet, what did I do wrong, right? If my dog got cancer. Is it genetic? I hear, dogs in Europe don’t get cancer. I spayed my dog, I didn’t spay my dog, I neutered my, like, what does all that mean? And before we haven’t had the data to do it, the other is we haven’t had the computational power to really take millions of data points. Right? And analyze them in an efficient and quick way. That’s part of what has been holding us up.

[00:16:33] The other thing we did recently that we’re going to share with people is we have the genetic data on these dogs. That’s like the whole DNA, right? We talked about the gene, genome.

[00:16:41] >> James Jacobson: Sequenced. We’ve spoken to the director of the NIH and comparative oncology. And, and yeah, this is pretty interesting. So you have, you’ve sequenced it. You know it.

[00:16:50] >> Dr. Kelly Diehl: We’ve sequenced all the dogs now. And which was really a huge.

[00:16:54] >> James Jacobson: How many?

[00:16:55] >> Dr. Kelly Diehl: 3000.

[00:16:56] >> James Jacobson: The entire population of the Golden Retriever Study.

[00:16:59] >> Dr. Kelly Diehl: Yeah.

[00:17:00] >> James Jacobson: They didn’t know they were getting into that when they got into that.

[00:17:03] >> Dr. Kelly Diehl: Oh, oh, heck no. Right.

[00:17:04] >> James Jacobson: Give me some fur. It started with like, send me some fur. Yeah.

[00:17:07] >> Dr. Kelly Diehl: Right. Well, and you know, we knew one thing that was really good about the people who set up the study is sometimes they’re like, we think hair is important or, you know, there’s, we know DNA is important, but in 2012, when they started collecting, the dog genome had only been the first one sequenced in like 2005.

[00:17:26] It was super expensive, right? Like, we can get our genomes done for a thousand dollars and the first one costs, what, a couple million, to get a human genome done. And same with dogs, but they said, you know what? Let’s save it. My favorite as a gastroenterologist is the poop samples, right? Because someone was like Well, let’s just save that poop.

[00:17:46] And now we know so much about the microbiome or we’re really interested in it. They didn’t know it in 2012, but they’re like, we think it’s important. And I give kudos to those folks who said, we think, we don’t know, but we think this is going to be important. So let’s save as much as we can.

[00:18:03] >> James Jacobson: It’s in freezers somewhere lovely in Colorado or where does all this poop live?

[00:18:08] >> Dr. Kelly Diehl: No, we actually, uh, freezers in Indiana.

[00:18:12] >> James Jacobson: Okay, okay. That’s good. You don’t have to, uh, but that’s amazing. So, so having all of this data that can now be analyzed by artificial intelligence is just sounds like, you know, to me, a game changer because you can look at these patterns that no one would have thunk of before.

[00:18:31] >> Dr. Kelly Diehl: Right. And just getting at some of the myth busting or saying like, this looks important over here because now we’ve got 100,000 dogs, let’s say. We got Golden Retrievers from here. We got dogs from the UK. There are databases in Europe because a lot of European animals have insurance. They’re covered under insurance.

[00:18:51] So there’s like a Swedish database and an Italian dog and cat database that are like a lot of big databases, right? We have a lot of insurance in the United States, it’s becoming more common. Those guys are actually now publishing a little bit of their findings. So we have the ability now to mash them together, right, in what they call meta analysis. And look at the big data.

[00:19:15] And then that can help point us maybe in a direction, which we need to look at more closely, right? And then you’re also bringing in artificial intelligence once you have the data to train it, right? Like, hey, dogs with lymphoma.

[00:19:33] >> James Jacobson: What, because without that, you just have a lot of data and then you’re relying on, like, someone to think, well, let’s look at this thing and this thing, and then, and really study it and spend a lot of time, which may be fruitless.

[00:19:45] >> Dr. Kelly Diehl: Right, exactly. A good example is, I think all of us have heard of the Framingham Heart Study, right? It’s in its fourth or fifth generation.

[00:19:51] >> James Jacobson: Well, for the listeners who aren’t familiar with it.

[00:19:54] >> Dr. Kelly Diehl: So, the Framingham Heart Study was started in 1948, and it was to look for risk factors for heart disease. Framingham is a town in Massachusetts. They’ve been following people, they follow their diet, their lifestyle for generations. I think they’re on their fourth or fifth generation of people. So it’s enormously important study. And what it did, it’s, it identified risk factors for heart disease we take for granted, but we’re not sure in 1948. For example, smoking or high cholesterol, or, you know, smoking and exercise.

[00:20:26] Before that, people, I mean, people weren’t dumb. They kind of had an idea. I think these go together. But there were like, for example, people thought it was normal for your blood pressure to go up when you got older. Like that was just some people’s blood pressure went up and that kind of happened. Well, we know that that’s not normal and it’s linked to heart disease.

[00:20:45] So Framingham was one of the first places to kind of go, hey, don’t pay attention to this. Pay attention to, for example, cholesterol. And we know the cholesterol story is super complicated, right? Because first there’s cholesterol. Then we were like, oh, there’s good cholesterol and bad cholesterol.

[00:21:01] >> James Jacobson: Good cholesterol, bad cholesterol.

[00:21:02] >> Dr. Kelly Diehl: Right. And it morphed from there. But without Framingham, we might have totally ignored cholesterol completely. That’s what we are wondering with our dogs as we complete the study. Is maybe there’s a factor we’ve just been blowing off or we haven’t recognized as, you know, this is important and this isn’t, so focus on this. Go look at this harder and that’s part of the reason for these big longitudinal studies like the Golden Retriever Lifetime Study.

[00:21:33] >> James Jacobson: You mentioned the concept of myth busting. So, you know, there are a lot of myths or a lot of concerns when someone is dealing with dog cancer. Is it the food? Is it, you know, that I spayed or neutered too early or didn’t or whatever? And so the results from analyzing this data can lead to some real answers as opposed to conjecture.

[00:21:54] >> Dr. Kelly Diehl: Yeah, we’re really hopeful. We’ve seen some trends with hemangiosarcoma. We published this last June with spay and neuter. Males and females were different. Males look like it didn’t matter whether you were neutered or intact as far as incidents of hemangiosarcoma. In females it did, but the problem was we didn’t have that many intact female dogs, you know what I mean?

[00:22:21] So there’s only a couple that lived long enough or didn’t have cancer. And we didn’t parse out in that particular study. It was more descriptive. Like, if a dog was spayed and developed hemangiosarcoma, well, were they spayed at six months or one year or one and a half year? Like, does that make a difference?

[00:22:38] >> James Jacobson: And you’re looking at just hemangio, whereas mammary cancer or other cancers could have different results based on that.

[00:22:44] >> Dr. Kelly Diehl: Exactly. And we’re going to do that. Here’s a little insider information. We’re going to probably do it with lymphoma next, because we have a lot of lymphoma. But that’s where we’re going to go. And we’re going to try, we’re trying to dangle the, like cat toy in front of some of our researchers and go, look at this intriguing hemangiosarcoma. Would you like to take a bigger, a dive, deeper dive in this, please? And we have more dogs, unfortunately, that even since that study have died of hemangiosarcoma. So we have even more data to look at.

[00:23:18] >> Molly Jacobson: Healing begins with hope, and if you’re facing dog cancer, you need hope. Real hope, not false hope. For real hope, visit our sponsor, DogCancer.com, where you can get better information today, so you have no regrets tomorrow. At DogCancer.com, we understand the anxiety, and urgency that comes with a dog cancer diagnosis.

[00:23:43] We’re dog lovers first and foremost. What are we second? Well, we are Team Dog, a team that includes veterinarians, veterinary oncologists, cancer researchers, and science writers. And we’re all here for you. So don’t wait, visit DogCancer.com today to learn more, find support. and join a community that really cares about what you and your dog are going through.

[00:24:05] And now, let’s go back to James Jacobson’s conversation with the ever hopeful Dr. Kelly Diehl of Morris Animal Foundation.

[00:24:14] >> James Jacobson: Are you guys ever going to do, like, does the Lifetime Golden Retriever Study end at a certain point, or is it just going to continue on?

[00:24:23] >> Dr. Kelly Diehl: Good question. It was originally designed to end this year in 2024. And, um, so I have to give a little lecture on statistical power. So what we decided was we needed 500 diagnoses of certain cancers, right, to adequately power the study. And what that means is 500 diagnoses is enough to make conclusions, right? Or at least make some pretty strong recommendations or pick up a risk factor.

[00:24:49] We anticipated those 500 diagnoses would be reached this year, 2024. End of study. Like, that was gonna be the end. We reached the 500 diagnoses last year, actually, which shows how many dogs get cancer who are Goldens, sadly. And it was decided probably a couple years ago. We’re full steam ahead till all the dogs pass. Or drop out or whatever. So we got a few more years to go, which is awesome. So we’re gonna go for the lifetime which was, we did not anticipate.

[00:25:22] >> James Jacobson: Wow. How many more from the original, how many more are still with us? How many senior dogs do we have?

[00:25:28] >> Dr. Kelly Diehl: The average age is a little over 10 in the guys in the study. And we are, unfortunately, we just hit like the 1500 mark. So half the dogs are gone, probably a little more than that. Not all are deceased. Some of those are withdrawn for some reason or what we consider inactive. But our inactive dogs can jump back in. Sometimes people move, things happen, life happens. We miss a year, we miss two years.

[00:25:55] >> James Jacobson: It’s a lot of work for the dog lovers in terms of visits to the vet and all that stuff, right?

[00:25:59] >> Dr. Kelly Diehl: Yeah, it’s really, um, it’s a lot. They’re a dedicated group, I will tell you that. They are a very dedicated group.

[00:26:06] >> James Jacobson: They have their own Facebook group, they are very dedicated. We’ve spoken to some of them and it’s an amazing group of the people who are associated with these heroic dogs.

[00:26:15] >> Dr. Kelly Diehl: Right, right. But it’s a big lift for us. It’s a 32, 33 million dollar study. And when you think over 75 years, we’ve given 160 million away in grants, that 33 will be a big chunk, right? For us, it was a big commitment. I won’t even tell you what we estimated at the beginning. It was way off. And then we had to increase it. It’s been stable for It’s like 10 years now, but at first we were like, Oh, it’s only gonna, I won’t even say. It’s going to cost this. Oh, maybe it’s going to cost us because the other thing is you’re storing samples for a long time and people are.

[00:26:53] >> James Jacobson: Those freezers in Indiana. Yeah.

[00:26:55] >> Dr. Kelly Diehl: These freezers in Indiana cost quite a bit of money. And the other thing we found that we didn’t anticipate is with technology improvements, people go, Oh, I don’t need that whole ml of blood. I only need half of it or quarter of it to do my studies. So that’s great. In some ways, more sample, like it goes longer than originally anticipated, but we kept a lot of samples thinking people were going to need large volumes.

[00:27:22] So we have to keep those samples, right? And they’re, they’re not. inexpensive. Some of them are easy to keep. You put them in a refrigerator basically. Some of them have to be very precisely stored with very temperature controlled conditions or they degrade. And those are the pricey ones.

[00:27:41] >> James Jacobson: So all the funds that you guys dispense are raised privately, right?

[00:27:48] >> Dr. Kelly Diehl: Pretty much. A huge chunk of it. We have some corporate partners that have given us money over the years. Like, Golden Retriever Foundation, for example, which is not corporate, but we have breed clubs like Golden Retriever Foundation provided a lot of money for us. The Morris Family personally chunked up some money to get the study going.

[00:28:07] We have, um, you know, Hill’s, which Purina, Elanco, but the vast majority for this study comes from private donations, which I think is also really important because you want to maintain scientific integrity. Right? Because that’s always called into question, like, Oh, you got money from so and so. Are they directing you? And the answer is a firm, no. We are independent. They may ask for samples or data. But they don’t control like how we collect it or who we give it to or what we publish, even if it can be controversial.

[00:28:48] >> James Jacobson: Let’s talk about the sharing of data because there are all these different entities now doing it and you know, you look at one medicine and you look at what you’re doing with animals and you look at all these different things. How collegial is the, you know, ecosphere where the scientists share this data? Is this something that, like, is easy to get if someone else is doing a study, I mean, in the science community? Or is it a little territorial?

[00:29:14] >> Dr. Kelly Diehl: Uh, a little bit of both. Um, that’s why we decided to to do the genetic sequencing ourselves so that someone wouldn’t take it, sequence it and say, that’s mine. That’s mine to look at. By doing it ourselves. And by ourselves, I mean, we had a researcher, we paid them to do it.

[00:29:35] >> James Jacobson: But you own the IP?

[00:29:36] >> Dr. Kelly Diehl: We own it. Yeah. And so we can share it and be like, all people have to do is apply to us so that we know kind of where it’s going. But then it’s like free to share. And I, I am really excited about that.

[00:29:49] The other thing we see with a lot of our grants is people don’t have enough cases, let’s say, at University of Florida. So they’ll often bond together and my experience compared to human medicine is that I think it’s more collegial. It’s a smaller group. People have personal relationships with each other and I see a lot of different, different institutions maybe are focused in different areas are like, well, I don’t want to compete with you.

[00:30:18] I’m going to focus on this and I’m going to focus on this. And I do see some sharing. And like I said, a lot of collaboration between institutions, like just to get case numbers, for example, plus studies are expensive. Like they get money from us, but we’re not the only, if, if you look at a study that’s published, you’re going to see Morris Animal Foundation, maybe an NIH grant in there, right?

[00:30:42] Other groups that have funded studies. So also by bonding together, sometimes you get a little bit more bang for your buck, right? You’re bringing more assets and resources into the mix.

[00:30:54] >> Molly Jacobson: Let’s pause here for a word from our sponsors and then we’ll be right back to talk more about international studies.

[00:31:04] And we’re back with James Jacobson’s conversation with Dr. Kelly Diehl of the Morris Animal Foundation.

[00:31:11] >> James Jacobson: So NIH is a part of some of these, which is U. S. What about international government agencies that study health?

[00:31:18] >> Dr. Kelly Diehl: That’s a good question. That’s harder for me to answer, but I know that some of our international grantees are getting money from their governments, right? So again, they’re cobbling funds together because we may give out average study now costs well over a hundred thousand dollars for a two or three year study, right? So, I’ve seen in my ten years here, things go from like 75, 000 to 150, 000 easy, right, for a two year study.

[00:31:44] >> James Jacobson: And that’s how many dogs, how many dogs go in a hundred grand study?

[00:31:47] >> Dr. Kelly Diehl: It depends, right? Sometimes that’s just a bench research. That’s a two year in the test tube, maybe, stuff. But what we’ll see is that people will get funds. It’s really a $500,000 study. We’re ponying up this much. Maybe they got some from their institution, maybe their government, you know, they applied for government funding.

[00:32:06] Maybe they got a postdoc that got a grant so you don’t have to pay that person’s salary. And all of them go together. All of that goes to publish maybe one, one result. So it’s really, it’s not just like paying for the test tubes, but sometimes you’re paying overhead, right? Because your lab costs you some space. You may be helping to pay for a technician who is on that study or a graduate student, right? Graduate students, you got to pay them sometimes and postdocs.

[00:32:36] And so it gets really complicated. We actually do very close examination to truly have the budget because we also feel like, okay, you’re giving me money. Like I want to be the best steward of that money. How are these people spending the money? Are they buying an expensive machine? Well, maybe that’s not a great use. Is it all salary? Well, maybe we don’t want to pay all salary. So we scrutinize all those little pieces that go into a grant when we fund something.

[00:33:07] >> James Jacobson: And you said you’re getting more funding requests from places that you never would have considered, or scientists who you probably, you know, when they saw the word animal, they were like, no, that’s not, that isn’t a good source. That must make your selection process a little bit more exciting because you, you have such a vast number of different things that you could potentially pull your money to.

[00:33:28] >> Dr. Kelly Diehl: It is, but we were just talking this morning about, you know, falling short. We always fall short. We always have more grants that we recommend for funding than we have money. We just went through a wildlife call where 33 grants that were recommended for funding, we can’t fund. So that means those grants were worthy of funding and we have no money.

[00:33:50] We’ve, you know, we rank them and we ran out, but there was nothing wrong with the grant. Like they were great grants. And we find that, you know, we were talking about the canine cancer, you know, there’s always grants recommended for funding that we don’t have money. Typically we might recommend, um, we have an independent board that comes in. So it’s not even us, right? We get a group of experts, they discuss amongst themselves. We just say, here are the grants, tell us what you think.

[00:34:17] Because again, it’s trying to be independent to get the best people to look at these things, not us, and they’ll recommend and they’ll rank. And typically we might say 15 to 18 are worthy of funding like these are great and we’ll have, they’ll be screening 40 to 50 at a time and those have already been screened but of that 15, 6, 7, we’ve got enough money for, that’s it, and our development officers will scurry off and they’ll often if they have a donor, but it’s not, most people can’t afford a hundred thousand, you know, to be the donor sponsor of a study.

[00:34:57] So sometimes we’ll pull money from other places if we have leftover. We’ll say, well, we, we didn’t fund all of our cat fellowships, let’s say. So maybe we can pull the money over here, but we still fall short all the time. And these are worthy grants. So then those researchers get a chance to apply again, but that’s a lot of work to keep applying over and over for grants.

[00:35:22] And in the meantime, you have new people applying for grants. So it is, um, we have a huge, huge need. and really good stuff out there and we fall short and we’re big, right? We’re a big organization and we fall short.

[00:35:38] >> James Jacobson: You’re the largest actually when it comes to this, right?

[00:35:41] >> Dr. Kelly Diehl: Yes. We say we are probably the largest other than, you know, some universities probably give are considered nonprofits, right? And they give grants to their faculty and you could say maybe they give Harvard probably gives more, you know, overall, but yeah, we’re the biggest animal focused funding agency for, you know, pets, horses. USDA will fund the right livestock, but.

[00:36:05] >> James Jacobson: But companion animals.

[00:36:06] >> Dr. Kelly Diehl: Yeah, nothing for companion.

[00:36:08] >> James Jacobson: So when you look back on all the work you’ve done, and again, to people who are interested in Dog Cancer Answers, what are some of the most tangible outcomes of the research that benefit dog lovers who are dealing with dog cancer?

[00:36:23] >> Dr. Kelly Diehl: Yeah, that’s always tough. I think we have some work, I’m going to go backwards and move forwards, we did some work on feline leukemia, and I, I’m going to say that even though it’s dog cancer, because by finding feline leukemia virus, which is a big driver of cat cancers and the vaccine, there’s what they call the before times and the after times with that. So that was one thing.

[00:36:49] As far as dog cancers, some of our work looking at different chemotherapeutics, in the late 80s, early 90s, some of which, like, it didn’t work, but that’s important to know, right?

[00:37:02] >> James Jacobson: You gotta know what doesn’t work as well, yeah.

[00:37:04] >> Dr. Kelly Diehl: Right. These drugs fell short. We did some work on, people probably know, Palladia, which is used on a lot of times for mast cell tumors. We did some work with that, some independent work, after it was out on the market. We’ve got, I think just the things we’re finding with the Golden Retriever Lifetime Study, like some of the stuff about, you know, this early work on spay neuter and hemangiosarcoma. I think that’s interesting. And we need to explore more on it.

[00:37:34] Some of the diagnostic stuff again, sometimes I hate to say it, it like didn’t work, you know, which is hard to say to people. But it’s important to know, like, there’s been stuff where people were really excited, like, we think we found something, uh, you know, we, we got it in three dogs and now we need money to do it in 50 dogs. But in 50 dogs, it doesn’t work.

[00:37:56] >> James Jacobson: Right.

[00:37:57] >> Dr. Kelly Diehl: We did something with osteosarcoma or recently we just finished the osteosarcoma vaccine. It was done at University of Pennsylvania. It has some very interesting findings and it may really help dogs with osteosarcoma, particularly those that maybe can’t be amputated, can’t have a limb amputated and actually that work, it’s not for dog lovers, but it’s actually informing a new clinical trial in kids.

[00:38:21] >> James Jacobson: Which is really the, I mean, kids are the only people who really get osteosarcoma.

[00:38:28] >> Dr. Kelly Diehl: It’s similar to dog.

[00:38:30] >> James Jacobson: Right, right.

[00:38:30] >> Dr. Kelly Diehl: That’s similar to dogs. Right, right..

[00:38:31] >> James Jacobson: Osteosarcoma is a lot more prevalent in dogs than in humans, and humans who get it tend to be children.

[00:38:35] >> Dr. Kelly Diehl: Right, right. So I think some of the, we did some diagnostic work with looking at different types of lymphoma that’s now pretty commonplace. We funded some of the research. on that, but we’d like a big win. We’d like to find a big win right now. And we’re hoping that the push that we’re doing with hemangiosarcoma and really the gamble we took with the Golden Retriever Lifetime Study, right? A big gamble will actually finally move the needle on some of it because it hasn’t moved in 30 years.

[00:39:08] Really, some of these, I don’t need to tell people out there. I had a dog with lymphoma a few years ago, and I’m like, this is the stuff I was using back in 1992.

[00:39:18] >> James Jacobson: Right.

[00:39:19] >> Dr. Kelly Diehl: And this is 2012 and we had zero, you know, like advancements.

[00:39:24] >> James Jacobson: So to turn this back to AI and this data mining, and obviously what you’ve basically expressed is that science is slow. And sometimes we were optimistic, but it doesn’t pan out. But what you have been gathering and storing in Indiana and places like that is all of this data that, I wonder, do you guys see an opportunity to go back and look at some of the things that you’ve considered in the past and reconsider them, but instead of having, you know, your team of researchers, you know, giving it to Hal, the AI?

[00:39:58] >> Dr. Kelly Diehl: For sure. And that’s where I think some of these big data sets can provide some information. And again, we have to train that AI, right? We can’t just.

[00:40:08] >> James Jacobson: You can’t just give it to her. But yeah.

[00:40:10] >> Dr. Kelly Diehl: Right, right. So, so, but how do we train them? Like, what do we tell them to look for? Sometimes we don’t even know, right? What patterns do we train the AI to say, here’s what lymphoma looks like? The other thing that we are really excited about and has never been done before is to have sequential blood samples in dogs that have developed cancer, right? So we have them from when they were six months old To let’s say eight years old and they develop cancer and it’s all the time and we have a lot of people looking at that, right?

[00:40:44] They’re like, pull me a hundred dogs that have developed lymphoma. The other thing we can provide is controls, right, ’cause you gotta have those controls. Like I find something, is it in the controlled dogs, age matched, sex, everything. That’s, I think something else we don’t talk about is we can provide controls for people that you may not have at all.

[00:41:07] And we can pull those same time. And we send them off to a researcher. And we have several researchers who are looking at, hate to use the term liquid biopsy because sometimes it gets it bad rep, but that’s it. Right? Are there biomarkers in the blood way before we see anything that we just are missing and that we can see. So there’s patterns, which is AI, right? AI can look for patterns, but there’s also, yeah, looking for biomarkers.

[00:41:36] >> James Jacobson: But I mean, I’m wondering if you’re applying the AI to kind of check the results that you’ve seen in the past to like, ’cause that was very human biased. And you know, what I’ve discovered in AI in general is that you can, if you train it well and you ask it the right questions, you can get it to critique what you did and perhaps see new insights and new patterns that the person and the people, because these are not, this is not large numbers of people, may have missed or not thought of or considered. Is that part of what you’re looking at?

[00:42:06] >> Dr. Kelly Diehl: Not yet.

[00:42:07] >> James Jacobson: Okay.

[00:42:07] >> Dr. Kelly Diehl: So honestly, no, not yet. I think that will come as we have AI programs that are like, Okay, we can, we’ll detect these patterns.

[00:42:19] >> James Jacobson: Or review your research, yeah.

[00:42:20] >> Dr. Kelly Diehl: Right, right, right. But that would be really exciting, right? Because it’s all about early detection or. and seeing patterns that maybe we don’t recognize like, Oh, this red cell count changed a little bit, but it’s within normal range. Right? But that it’s not, it’s significant.

[00:42:39] >> James Jacobson: Because it’s readily, we accept it, you know, standard practice.

[00:42:42] >> Dr. Kelly Diehl: Right.

[00:42:43] >> James Jacobson: We accept that that’s the normal range, but normal ranges in different countries varies in different.

[00:42:48] >> Dr. Kelly Diehl: Right or maybe those changes, those subtle changes, and that’s what you know looping back to that, the stuff I told you about the AI and the hyperadrenocorticism is the AI was picking up like just these little changes, but taken as a collective, right, a pattern. It would go, you should look at this dog. And I’ll go back to a study I saw out of Davis where they looked at the flip side of Cushing’s, which is Addison’s disease, hyperadrenocorticism, notoriously difficult, difficult disease to detect. And at Davis, they started some work with, let’s see if we can train the AI, let’s do AI, see if they can pick up.

[00:43:30] And we got all these dogs coming into the referral for, you know, their referral hospital. And so they trained the AI. They said, here’s what a bad case of, of hyperadrenocorticism. These are all confirmed cases. Look what they look like. Here’s all the data. Here’s all the blood work. Here’s the history, all of this. And then they said, okay, AI, they sicked the AI on every dog that came in the hospital, whether it was coming in for a broken toenail, broken bone, suspicion for Addison’s disease, heart disease, whatever.

[00:44:01] And the AI would go, I think you should go look at this dog. And I was in a section, a lecture where they put this up and this is a group of internists and they said which of these dogs, they put up the blood work and the history, they said, which one of these dogs do you think has Addison’s? And we’re like, none of them, like, would you do testing on any of these?

[00:44:21] And one of them that, like, they found Addison’s in a dog that came in for, like, a hip surgery. Like, there is no way, just based on the patterns. And they tested and they’re like, my gosh, this dog had this. So suppose we could get to that point with cancer. Like, these little changes you don’t think are significant, we’re wrapping a computer brain around and guess what? We can recognize a pattern and you need to go look.

[00:44:50] You need to start looking in this dog or watching them at least, right? Because that’s the other piece of AI that I’ve heard from folks is, okay, so we find this, we think we found something early, we think we found hemangiosarcoma, what do we do? Like, do we give them, hit them hard with chemo?

[00:45:08] >> James Jacobson: How invasive? Yeah.

[00:45:09] >> Dr. Kelly Diehl: Right. Is it invasive? Do we take their spleen out? Do we, you know what I mean, like, and, and some of the people I’ve heard.

[00:45:15] >> James Jacobson: And maybe this is where the AI enhanced imaging or analysis can come in and like.

[00:45:20] >> Dr. Kelly Diehl: We would hope. But it’s the one thing I’ve heard, not controversy, but people push back on. Okay. So we find this. But what do we do? Like, do we do something invasive? Because then the next step is we’re going to start giving dogs chemo, right? We’re going to decide, is it prolong their life? Does it make a difference if we pick it up at such a point?

[00:45:43] And you know, intuitively we might say yes, but I think we know there are cancers in people that they will sit and watch, right, because there’s no benefit to giving you like, um, there’s certain forms of lymphoma that are very slow, right? And I have a friend who has it and she’s not had chemo. She’s had cancer for seven years and hasn’t had chemo because they’re like there’s no benefit. Like we don’t want to make you sick.

[00:46:10] >> James Jacobson: The treatment is worse than the long term, you can probably die of something else.

[00:46:15] >> Dr. Kelly Diehl: Right. So do we have dogs like that that’s gonna be the next step. But the first step is finding them in the first place.

[00:46:20] >> James Jacobson: Yeah.

[00:46:21] >> Dr. Kelly Diehl: Right? And then we can decide what do we do ethically from here?

[00:46:25] >> James Jacobson: Okay, so polish your crystal ball and tell me how from what you see from your vantage point where AI can be headed in terms of cancer and dogs.

[00:46:35] >> Dr. Kelly Diehl: Picking up patterns early. And that entails us taking our pets, and I am going to raise my hand as not always great at doing that, especially if you have difficult pets that hate the vet, right? But without that data, other than maybe lifestyle, I’m going to polish my crystal ball and say, there may be some lifestyle risk factors. And that’s the other thing. Those could be modified, right?

[00:47:00] The other thing we’re looking for is modifiable risk factors. So, if AI can say, don’t do this, then you’re going to tell people who come in with their puppy, you know, just like we tell them, do this dewormer, or, or you should do these vaccines or whatever, or this is how we feed large breed dogs versus small breed dogs, whatever, that we can tell people right off, right?

[00:47:21] These are risk factors, like, you know, we know smoking’s bad. We know secondhand smoke is bad for our animals, right, and has been associated, you know, it’s a little shaky, but still with cancer, especially in maybe cats. And pesticides, herbicides on lawns, right, there’s some evidence that exposure to that is a problem.

[00:47:41] But we need more data that says, yeah, yeah, absolutely, you shouldn’t do this. And that’s what I’m hoping AI can tell us is it can pick up a pattern looking at lifestyle, blood work, all these factors, diet, spay, neuter, and smash them together and say, here’s what I recommend. And I am also going to stick my neck out and say it will be different for every breed of dog and size.

[00:48:05] >> James Jacobson: Wow.

[00:48:06] >> Dr. Kelly Diehl: There will be different recommendations for different dogs. So you heard it here first, yeah.

[00:48:11] >> James Jacobson: Okay. That’s cool. Well, that is a great place to end Dr. Kelly Diehl. Thank you so much for being with us today. We, we think the work that you’re doing is amazing and we’re going to continue to cover this in some of the individual studies, especially when it comes to AI, because I think it’s fascinating.

[00:48:27] >> Dr. Kelly Diehl: Yeah, for sure. So we’ll have to stay in touch, and I’ll let you know how those AI ones and the big data ones come out, right? So thanks so much for having me on. It was so much fun.

[00:48:39] >> Molly Jacobson: And thank you, listener, for being here. Please remember to check the show notes for relevant links and follow us wherever you get your podcasts to hear new episodes as soon as they drop.

[00:48:49] If you’ve got a story of hope and healing or a question for our producers to explore, go to DogCancer.com/ask and let us know.

[00:48:58] Until next time, I’m Molly Jacobson. And on behalf of all of us here at Dog Podcast Network, I’m wishing you and your dog a very warm Aloha.

[00:49:12] >> Announcer: Thank you for listening to Dog Cancer Answers. If you’d like to connect, please visit our website at DogCancer.com or call our listener line at (808) 868-3200. And here’s a friendly reminder that you probably already know, this podcast is provided for informational and educational purposes only. It’s not meant to take the place of the advice you receive from your dog’s veterinarian.

[00:49:35] Only veterinarians who examine your dog can give you veterinary advice or diagnose your dog’s medical condition. Your reliance on the the information you hear on this podcast is solely at your own risk. If your dog has a specific health problem, contact your veterinarian. Also, please keep in mind that veterinary information can change rapidly, therefore, some information may be out of date.

[00:49:55] Dog Cancer Answers is a presentation of Maui Media in association with Dog Podcast Network.