Could your dog save another dog’s life just by taking a sniff? The answer is a resounding yes! Join us as we dive into the fascinating world of biomedical olfaction with Dr. Cindy Otto and Dr. Clara Wilson, authors of the groundbreaking paper, Trained Dogs Can Detect the Odor of Hemangiosarcoma in Canine Blood Samples. Hemangiosarcoma—a silent killer that devastates breeds like Labradors, German Shepherds, and Golden Retrievers—is often diagnosed too late. But pioneering work at the Penn Vet Working Dog Center is changing that. Discover how these incredible canines are using their powerful sense of smell to accurately detect the deadly cancer in blood samples, achieving over 70% accuracy and even revealing discrepancies that stumped traditional pathologists. Learn about the intensive training that turns working dogs into life-saving disease detectors, and what this research means for the future of veterinary medicine and early cancer screening.
◇ ─ ◇ ─ ◇
Froggy’s Golf Ball Retrieval offers the best programs helping you achieve the most from your water hazards. Not only will we treat your course with the utmost professionalism, our experience in removing balls from water hazards has taught us how to operate in an efficient, safe and productive manner since 1995. As we move into our next chapter, Froggy’s provides insured services, uses proven techniques and the most updated equipment to take care of the golf community.
We handle even the toughest water hazards that make our competitors walk away. Froggy’s is the Midwest’s premier ball collection company. Froggy’s offers several options to make your water hazards a profit center!
Please call us today to arrange to have Froggy’s harvest the balls from your water hazards.
Contact Owner Emily Newland at 574-544-9890 or email Froggysretrieval@gmail.com.
◇ ─ ◇ ─ ◇
All footage is owned by SLA Video Productions.
—
Watch the episode here
Listen to the podcast here
Training Dogs To Sniff Out Cancer! – Dr. Cynthia Otto & Dr. Clara Wilson
Training Dogs To Detect Cancer In Canine Blood Samples
Thanks for tuning in. Janet, I am so excited about meeting our guests. Dr. Cindy Otto and Dr. Clara Wilson are both the authors of Trained Dogs Can Detect the Odor of Hemangiosarcoma in Canine Blood Samples. What all that means is they are training dogs to detect cancer in other dogs. I love that because dogs can detect cancer in us. I want dogs to live longer. I want to find out about cancer early in dogs, too, so we can do something about it. Welcome to the show, Dr. Cindy Otto and Dr. Clara Wilson. Can you tell us a little bit about yourselves and how you came to be doing what you’re doing?
We have a Buckeye over there. I know it.
Go Bucks.
We probably saw each other in the halls because you were an ‘86 grad and I was ‘89.
I was probably there.
I was not in those hallways. I graduated in ‘81. I was in the business school.
We didn’t know that existed. The vet school is all its own little island.
Penn Vet Working Dog Center Opens For Research And Training
I can start because I’m older. I graduated from the Ohio State University in 1986, and then I went on and did an internship at the University of Pennsylvania. I did a residency at the University of Georgia and a PhD in Veterinary Physiology. I then came back to the University of Pennsylvania. I achieved my board certification in emergency and critical care, a little bit of an alternate route, and then had a research lab. I worked in the emergency room. Eventually, I got involved in search and rescue dogs. One thing led to another.
I deployed with our search and rescue task force that I had been a member of since about 1994. In 2001, it was that pivotal moment when we deployed to 9/11. In 2001, I deployed with the Pennsylvania Task Force to ground zero and took care of the working dogs, the search and rescue dogs, at that site. Following that, I was all in on search and rescue dogs and working dogs. I spent the next eleven years trying to figure out how to develop a research training and education center that supports these dogs and to understand them.
Also, because there was always a shortage of these dogs, how do we help address that? As part of that, I then became board-certified in canine sports medicine in order to support the physical well-being of these dogs. The big research that led through all of that was the American Kennel Club Canine Health Foundation-funded longitudinal study of the search and rescue dogs from 9/11. We’ve published about fifteen papers from the information we’ve collected from those amazing dogs. It inspired a lot of what we do.
In 2012, we opened the Penn Vet Working Dog Center as a research, training, and education facility. We have since had over 200 dogs come into the program. We’re close to our 200th graduate. We have a research basis because our real goal is to determine how we help support these dogs from a research standpoint. People like Dr. Clara Wilson, who’s a postdoc in our program, have helped push that forward.
We have published over 85 papers from the working dog center, sharing the knowledge that we have been accumulating. It’s an exciting program. It is so comprehensive and so unique because we bring in puppies at eight weeks of age. They come to work five days a week, and then they spend evenings and weekends with foster families so that they can become socialized.
They pick their own careers. They get to tell us whether they want to be a search and rescue dog, or maybe they prefer to be a police dog, or maybe they prefer to do some explosive detection work, or maybe be one of our research dogs to help with projects like the one that we’re talking about. This has been my life’s dream. It is my opus as well.
First of all, to be involved with the dogs of 9/11 and all they did for us. Now, to be doing what you’re doing with not only detecting cancer and everything else that you’re doing as far as the olfactory thing, but having dogs help each other, I love that. Dr. Wilson, tell us a little about you.
My entry into the field is a little different because growing up, I loved animals, and I had been around service dogs. My grandfather’s blind. I had grown up with dogs, and looking at that relationship in how dogs can help us and how we can help dogs. I was interested in that, but I didn’t want to be a vet. I was very interested in animal behavior and human-animal interaction.
I was lucky enough to go into psychology and have an amazing professor who specializes in animal cognition. As soon as I found out that that was its own field, I was like, “This is exactly what I want to do.” I pursued that then into a Master’s of Applied Animal Behavior and Welfare and then into my PhD. By that point, I was honing in on these different ways that animals can help us, but was especially interested in their olfaction. I felt like I had a pretty good sense of seeing eye dogs and hearing dogs, but I thought that was an underutilized world with that sense of smell that I was very interested in.
My PhD was all about canine olfaction and applied behavior. I was specifically looking at dogs being able to smell human psychological stress. When humans were getting stressed, their odor profile changed. I was able to show that the dogs can smell that when you take samples from people in the lab. They can discriminate between those samples. That was what my PhD work was in.
I’d been able to come over to the Penn Vet Working Dog Center a couple of times for different internships. I loved it and knew I wanted to come back as soon as possible. Once the postdoc opened up, I snapped it up because I couldn’t wait to come back. I’ve been here for almost four years, doing this biomedical olfaction work, but also a lot of canine cognition, puppy development, and some fitness, occupational health, and longevity work. I’m lucky to be able to dip into lots of these different research areas that Cindy set up.
What happens when you’re at a party, and someone says, “What do you do?” You say, “I’m an olfactory veterinarian.”
First of all, all they have to hear is veterinarian. All the words in front of it make no difference because then you get the questions like, “The dog had diarrhea yesterday, what can I do about it?”
They’re like, “They’re on some kind of medication.”
“It’s blue.”
I’ve got a question for you, too. You don’t only work with one particular breed, right? You work with a bunch of different breeds?
Identifying Highly Motivated Dogs For Olfactory Careers
We have preferred breeds. Part of that is because of the history of some of these breeds. We’re interested in genetics and what makes a good working dog. All dogs have the capacity to do olfactory tasks. We know that because with canine nose work, the sport, you can look at that, and every breed in the world is participating in that. They love it. It’s great for the dogs. If we’re looking for a dog that we want to do this as a career, we need a dog that has that motivation and the physical and mental capacity to do this repetitively. One of the challenges is to identify those dogs.
If we're looking for a dog to do this as a career, we need one with strong motivation and the physical and mental capacity to perform repetitive work. The challenge lies in identifying those dogs. Share on XMost of the dogs in our program are Labs or Shepherds. Whether that’s German Shepherds, Belgian Malinois, or Dutch Shepherds. We’ve had a few other breeds, and some of them are better suited, particularly in our environment. Our dogs come into this classroom five days a week, and they go home. They have lots of interns, volunteers, and other people interacting with them. Some dogs are not as flexible in their ability to handle the chaos. That’s what we were looking at.
In our program, our dogs graduate, and they leave us and go to jobs. There are some breeds that are not as easy to place in jobs because of people’s perceptions of those breeds. That’s another reason that we’ve mostly worked with the Shepherds and the Labs, but we’ve also been most successful. We understand those dogs. We understand how they learn. We understand how they adapt.
Ninety-three percent of the dogs that have entered our program have graduated into working careers, which is unheard of. We’re so excited. We think there are a lot of reasons for that. One being that we start so early. Another is that we don’t force the dog to pursue a career that is not consistent with what their needs are. We allow the dog to tell us, “Search and rescue is pretty cool, but I’m scared of heights. I’m not doing that.” They will go into a different career, but we still find careers that are successful and productive. I am so proud of that.
You may have noticed that in the paper that we’re talking about, there was a real mix of breeds. That’s because the paper was a little different in that we had one dog from our program, Vork, who is Cindy’s foster dog. Vork is a Black Labrador. She is a full-time research dog. Due to the scope of the project, and we had other projects going on that some of the other research dogs in our program were dedicated to.
This time, we opened it up a lot to what we call our community science program. That is either staff or people in the communities’ pet dogs who are allowed to come in and try for this type of research study. This specific study ended up having four community science dogs and one dog from our program, which is a little bit unusual. There, we do have this wider representation of breeds. We had a Hound mix and a Pitbull-Border Collie mix.
There is a bit more variability there, and that’s because the only job that these dogs are required to do is confined to the laboratory as opposed to maybe needing to go out onto rubble or into the environment. We do see sometimes a slightly wider net for these types of olfaction-based studies as opposed to some of the working roles, like a search and rescue dog, where you may need to be a certain size or something like that to be able to do the job.
I’m going to mention present company excluded right here. They did not mention Golden Retrievers.
I’ll let both doctors tell you about Golden Retrievers.
This one has exceptional sniffies.
We have a Golden in one of our research studies.
Dogs Accurately Identify Hemangiosarcoma And Inconclusive Samples
My question, and correct me if I’m wrong, is that the dogs were correct over 70% of the time in picking up the smell of the cancer in the blood sample. These were all confirmed to be hemangiosarcoma cases except for one, the trickster. I had to read that three times.
Do you mean it’s a fake one?
They’ll tell you. Would you tell Janet about that?
Sure. The way we got our samples was from various biobanks and also a lot from the Morris Animal Foundation’s Golden Retriever Lifetime Study. One of the prerequisites for coming into our sample base was that if it was a hemangiosarcoma sample, the dog had been confirmed through histopathology as having hemangiosarcoma.
We had over 400 in the study. We, by randomly using a number generator, pre-selected 36 to be in the actual test. Unfortunately, one of the hemangiosarcoma samples that we’d selected for the test, when we followed it up, wasn’t confirmed in the way that it should have been. There was some disagreement between pathologists.
The first had said, “This is hemangiosarcoma,” so it came to us as hemangiosarcoma. We didn’t know until we looked back at the numbers that the dogs did very poorly on this one sample. They were only around 30% something accurate. We followed up with the provider and said, “I can’t see anything in the spreadsheets to tell us that this sample is unusual, but the dogs didn’t seem to think it smelled like hemangiosarcoma.”
They came back to us and said, “Since it came to you, a second pathologist has said that’s not hemangiosarcoma.” They sent it to a third party. In the end, they decided it was too inconclusive to give it a yes or no, so they marked it as inconclusive, but it was clearly an unusual and not a cut-and-dry case.” That was very exciting because it suggests that the dogs were able to pick that up even better than the traditional methods.
That was an unusual one because it was a mistake, but it was a very helpful mistake. It told us that if you do have a case that is presenting like hemangiosarcoma but maybe isn’t, the odorous compounds are able to tell them apart, and they were pretty sure that it wasn’t. That was a slightly unusual piece that was an accident, but an interesting and helpful accident.
I thought that was impressive. I thought that these dogs are so good, and we’re not. Come to think of it, if we have something removed, I want 3 or 4 pathologists looking at it.
Alice, there is a dog named Otto in their program. It’s a female. I wonder who she’s named after.
It’s a male.
I thought that was weird because it says female.
I will have to fix that.
Maybe non-binary.
I think Otto’s pretty clearly male. He’s got a lot of testosterone. We had a kind donor who made a donation and was then able to name a dog. They chose the name Otto in my honor, which was a little embarrassing, but very touching.
That’s very special. I thought it was cute.
Does it make a difference in the olfactory senses if the dogs are spayed or neutered? Do you find that intact or neutered makes a difference?
We can’t answer that question because we don’t spay or neuter any of the dogs that are in the program. Our female dogs, once they hit about fourteen months, we will spay them if they’re not going to be part of a breeding program, but we don’t see any difference. Anecdotally, the only thing that I would say is that our intact males sometimes get a little distracted, especially because we have intact females who are often in heat around. They learn to work. They learn to figure it out.
Using Disease Control Groups To Ensure Specific Detection
I’m reading this study, and I’m like, “Yes.” They had dogs that had other diseases, too. That would be a ringer. I’m sure a dog with diabetes smells differently than one that’s hypothyroid. Something along my neck of the woods in Durham, or a Cushing’s dog, or a pseudomonas ear dog type thing. I commend you for doing that because I’m sure someone would have asked you about that.
It’s important. There are still some studies out there that don’t have what we would call a disease control group. We do need to take those with a pinch of salt because the dogs are good at finding differences, but we can’t ascribe that they have an understanding. I can say they’re smelling cancer if I only have a cancer group and a healthy group, but they might be smelling different, which could be all sorts of different things.
Dogs are excellent at detecting differences, but we can’t assume they understand what they’re smelling. Share on XMaybe they’re smelling inflammation or other metabolic changes that are meaning sick. That’s helpful and still interesting to know, but I then need to be very careful about not saying that they’re smelling cancer. That’s why we need that control group. We do see the dogs go through their different phases of understanding. As I show them more and more examples, they might initially alert on a diabetic dog or a dog with Cushing’s because they are using this wider spectrum of inflammation.
The more examples you show them and the more you teach them over time, “I’m going to reward you for alerting to this, but not this,” they narrow down. It’s interesting how narrow we can go. Here, we were looking at hemangiosarcoma. We haven’t asked these specific dogs what about other cancer types, for example. That’s the process that you need to do to make sure that you’re not just detecting different, but it isn’t necessarily your disease of interest.
I do want to make one comment, though. We often say we’re detecting dogs. We’re not having dogs sniffing dogs. It’s the same with human cancer. We’re not having dogs sniff people. It’s always samples. It’s always removed. It’s always in the lab. That’s a very different process than someone going out into the community with a dog and sniffing dogs, people, or whatever.
Look at the variables. You take a dog into someone’s home. They could have other pets. They could have cats. Smells galore. We know they detect it. How do you isolate those VOCs? I can’t even imagine how you would do that.
My question is, do we need to do that? To what end are we looking for those VOCs? There are ways to make this a diagnostic tool without knowing what we’re sensing. There are some electronic sensors that use the same agnostic approach. Certainly, our goal in working with the dogs is to identify what we can translate into a high-throughput system that can use something more mechanical to do the screening. The dogs are amazing, but it’s also labor-intensive. You think it’s inexpensive because it’s just a dog, but it’s expensive because of the training that goes into it. I’m going to go ahead and let Clara sort of address the VOCs in more detail.
The next question always is what are the VOCs. I think there’s still a bit of a gap. There’s this perceptual gap between knowing the individual chemicals and then what that perception is giving to the dog. We sometimes say, “I know the smell of Coca-Cola,” but it’s vanilla and lime. It’s these specific ingredients, but looking at those together, you’re not thinking of Coca-Cola because there’s this perceptual gap.
That’s where we’re at, where we are able to define VOCs. Some of our work is doing that. We’re working with chemists and engineers to have the same samples that the dogs did in that published study run through a machine to say, “We’re getting this amount of Chemical A and this amount of Chemical B.” The hard part is in translating that into a predictive model. That’s where dogs are still surpassing technology.
There are two potential reasons. One is this perceptual gap that I’ve described, where knowing X amount of these hundred different chemicals still doesn’t lead you to get this overall smell in the same way that dogs do. The other is that the machines still aren’t picking these very low volatiles up as well as the dogs. The chemists and the engineers know this. They are overtly open about the fact that their machines aren’t able to pick up some of these ones that the dogs we know can.
There’s a little bit of an issue there in that we’re missing some of the picture, but also knowing that each ingredient still doesn’t seem to necessarily translate to the ability of a machine to then say yes to this one and no to that one in the same way that the dogs are able to do. There’s a bit of a gap still, but that is what we’re interested in trying to make a machine that can replicate what the dogs are doing with that level of accuracy.
That’s amazing. When I was reading this, I thought that in the old movies and TV shows, you saw the bloodhounds going out. Now, we’ve got Labs, Golden Retrievers, and Shepherds.
No Goldens.
There are Goldens, but not in this specific study.
I’m thinking of these little Pugs and all the Frenchies that are out. Are any short-nosed breeds doing this at all, or is there a reason you couldn’t use a short-nosed breed?
I don’t think there’s a reason that you couldn’t. Dr. Otto will speak better on the veterinary side of things. My understanding from the literature is that there are some changes in the number of receptors and the number of genes that they have, but not enough to make them incapable of detecting the odor differences. There’s no reason why they couldn’t. I’m sure people are using brachycephalic dogs for this type of work. I haven’t personally come across one or had the pleasure of working with one, but I think it’s possible.
The other thing that you need to think about, though, is stamina. Where we may run into issues is the stamina piece. Although, it’s not hugely energy-taxing. They like to run in and out. Some of them like to go faster than others. We’re doing a lot of repeated trials. You need to be able to keep that sniff rate very high. I wonder if we may run into more issues with all of the extra soft tissue. There probably are some successful brachycephalic dogs, at least doing nose work and scent work. This is a little bit of an extra level-up because it’s not just the presence-absence question. It’s low saliency discriminations.
I know it was in the paper, but the 100 microliters that we were using in each box is a tiny drop of blood. It was a small amount of blood serum that they were seeing. I can’t detect anything. If I were to sniff along the line up, I’m getting nothing, let alone picking up changes between. The dogs do need to stay calm and not get too overheated. We know that once they get hot and are panting too much, it can interfere with their ability to do the task. We try to keep that to a minimum. They love the job. Sometimes they do get a little excited, and we need to take a break.
What does the training look like for these dogs? It starts very young, right?
Yes. Do you want to talk about the foundational training, or do you want me to talk about the lab training?
All of the dogs in our program are selected for their history of olfactory work. They’re either hunting lines or search and rescue lines. They’re already coming in with some good foundation. When they’re here, they start learning about using their noses from the time they’re eight weeks old. We use a synthetic chemical called the UDC, or the Universal Detection Compound. It’s 1-Bromooctane. That is a compound that used to be used in dry cleaning, and now is not. It is pretty much not anywhere in the environment. The folks at Florida International University developed it as almost like a calibration compound.
The nice thing about it is that we can work with our dogs and teach them to find this smell. It’s not going to interfere with any job that they do in the future, but it teaches them the mechanics of how to use their noses and how to search. All of our dogs are also trained to find live humans. Those two things happen from the time they get here.
Those pieces are telling us that the dogs in our program are using their noses. They like using their noses. If they don’t, then they’re maybe that very small sliver of dogs that will go on to do other things, like maybe one of our Golden Retrievers who left the program to be more of a comfort dog or a therapy dog. They all have that foundation. Even our community science dogs have gone through some training with this compound to make sure they’re up for the task. Clara, Molly, and our team get into the nitty-gritty of getting them into the very difficult task of finding these biological odors.
Training Mechanics Using UDC And Olfactometers
The UDC piece is teaching the mechanics. They need to be able to give us an alert. The alert is a change of behavior that tells me that they’re picking that individual sample. For the paradigm that we used for the paper, we were using these odor delivery devices called olfactometers. We also call them boxes because they look like a box. Each of those at a single time would be pumping the air from a single sample. Each one would have a drop of blood from one individual dog patient.
The nice thing about these olfactometers is that rather than me manually going in, and we have one box that is the hemangiosarcoma box, and I’m moving it around, and the dogs might be adding their own odor to it the more they pant on it, and I’m touching it, and all sorts of different contaminants can happen, instead, with the olfactometers, we have three in a lineup. The computer will switch on which valve is blowing the air. At a single time, one of them is presenting the hemangiosarcoma blood. The other one is a healthy dog’s blood, and the one next to it is a diseased dog’s blood.
Every time the dog is released to go smell the boxes, it will have moved around without anyone having to go in and touch anything. That’s been great for us to use that type of technology. The alert is still present. They would have learned how to do that with UDC. That’s a much simpler question because it might be 2 empty boxes and 1 with UDC. They need to tell us either by sitting in front of it or, more typically, standing and freezing over it that this is the box that they’re picking.
Are they wagging their tail and getting all excited?
We have a couple of dogs that have shaped different pieces. You know that they’ve finished their alert because their tails were doing this, and then it finally freezes, too. Other dogs are still in the front, but excited in the back. We have all sorts of different little spins on it. That alert tells me which one they’re picking.
When we move it to blood, it’s a lot harder because it’s not presence-absence. It’s not an empty hemangiosarcoma. It’s a healthy dog, a diseased dog, hemangiosarcoma. That’s a little different. It takes a lot of weeks to show them lots of examples. You start by showing the hemangiosarcoma and clicking or doing your marker word and rewarding them as soon as they sniff it. You build that out a little bit longer.
You also quickly need to involve those other samples, so they learn, “I need to discriminate between the two.” It’s a hard job for the dogs because they, at some point, are going to be wrong because they don’t know this category yet. We’re adding in category formation. That’s where I think it’s an interesting but unique job for the dogs. It’s very different from some of our other scent domains where it’s more of a, “Is this my target odor or not?”
Here, I don’t even know what about the hemangiosarcoma is their target. I have to keep showing them examples and saying, “Yes, it’s this. No, it’s not that.” Every week, I’m showing them a new dog so they can’t cling to their memory of what they saw last week. It’s an interesting task because it’s mixing that discrimination with this category formation where they’re having to learn this general odor that I can’t perceive myself. It does take a while.
What made you decide on hemangiosarcoma and not something like osteosarcoma or even lymphoma?
I have a doctor’s question. Have you ever created letters to say that word instead of having to say hemangiosarcoma? HMS or something? I don’t know. I’m making it up.
The initials we use are HSA. Autocorrect invariably turns it into HAS.
Every time you guys are saying it, I’m like, “Uh-huh.”
We also call it hemangio for short.
There you go.
The reason that we chose hemangio, there’s a method to our madness. We started our cancer detection with human ovarian cancer. Human ovarian cancer is a silent killer. There is no diagnostic test for it. If it’s diagnosed late, the prognosis is grim. Hemangiosarcoma or hemangio is the same in our dogs. Part of what clinched it for us, because we’ve been talking about it for a while, is that it is the number one cardiovascular cancer, particularly of working dogs.
In our study of the 9/11 dogs, of the dogs that developed cancer, whether they went to 9/11 or not. Hemangiosarcoma was universally the most common. The breeds that we tend to use for working dogs, such as Labradors, German Shepherds, and Golden Retrievers, are all at very high risk for hemangioma. We have lost several of the dogs in our program, and several of the partners of our working dog handlers historically have been through this.
Having worked in the emergency room for the first 20 years of my career, the devastation of having a dog that was fine two hours ago and collapsed, and you give them this horrible prognosis, there’s a whole lot of need. It’s a silent killer. It ticked all the boxes of, “If we’re going to do this, this is the one that we’re going to tackle first.”
I lost my first dog to hemangiosarcoma. I had another career. I wasn’t a veterinarian then. Whenever something happens to your first dog, you always follow that disease the rest of your life. To be honest, I don’t see that we’ve made that many strides in diagnosing it or treating it. That’s why I was so excited when I read your paper, because I thought if there was something we could do to pick that up early, it would be a godsend. Janet, I was walking down to the mailbox with Muncho. This was my first dog. We’re walking, and he falls over.
Muncho is short for Munchkin because he was this big.
He was a Hound mix from New Jersey. He fell over, and I went, “He’s dead.” After twenty seconds, he stood back up. I had gotten into vet school. I did everything I was told. We had the surgery. Three months after the surgery, he was peppered with cutaneous lesions. It was not good.
Is that what my Shaggy had?
Yours had a mast cell tumor on the digit.
Those, whatever they’re called, started popping up. You couldn’t stop it. It was so tragic. He was ten.
I’m curious. Do they ever have a bad day? If you guys aren’t feeling well, do they come around you like, “What’s wrong? What’s going on with you today?” Do you ever notice that?
All of my PhD work was on smelling psychological stress and proving that dogs could smell it. There’s almost a test effect. We have community science owners. The dog’s actual owner will come into the room, give the dog the treats, and be part of the process. We’re all blinded, so we don’t know where the sample is or which one’s right. On test days, the dog’s performance would sometimes drop even though nothing else had changed. We’d gotten them used to blind testing.
We were beginning to mess around with not even telling the owners that it was a test. It may be our body language. I can’t say that it’s for sure the olfaction signals. It might be that we’re nervous. The dogs pick up when it’s like, “This one’s important,” versus a normal training session. Everyone’s relaxed. We have tried to minimize the difference between a training and a test session so that the dogs don’t pick up on these differences and feel different.
In terms of having a bad day, they definitely do, which is why we would love to be able to move to an automated test for the scale-up. For now, they’re the gold standard, so we’ve got to keep working with them until we can figure out how they’re doing it. They do sometimes have off days. Either it precedes them, as maybe they’re not feeling very well. If there’s any kind of respiratory thing that they have going on, we know that impacts olfaction.
One of our dogs, we found out later, had a bad back, which we managed. When that was bothering him, it was like me being asked to do a complicated math exam when I have a bad headache. I might not be able to do it to the best of my ability. We do see some variability, but usually, it’s points in time. As the training’s going, we see a certain standard, and then if the dog’s not meeting that standard, we think, “Let’s see what’s going on today.” Usually, we can figure out why. Usually, it is tied to some of these health things or that kind of stuff.
Next Steps Defining The Hemangiosarcoma Signature And Creating A Mimic Odor
We’re working with chemists and engineers down at Texas Tech University who are amazing. They’ve been running the samples that we used in the published study through their machines. These are gas chromatography and those types of machines to pull out what’s in each sample. That’s been exciting to begin to get a picture of what the dogs could possibly be picking up on. Knowing the list doesn’t necessarily give you the perceptual piece, but it’s a step forward.
That would be the blueprint for defining which chemicals may be the hemangiosarcoma signature. The keyword there is signature because it’s not, unfortunately, going to be 1 or 2. It’s going to be the way that they talk to each other in combination. Still, this is the best way, if you want to look at specific chemicals, to move forward.
Once that’s been established, they’re working on creating something called a mimic or a pseudo odor. That means that they’re going to try to replicate what we found and create a soup or a magic potion. We’re then going to give that back to the dogs who are proficient in the real smell and see whether we can get anything that’s close enough that when they see it, they say, “This is what I’m used to detecting as hemangiosarcoma.” That’s an interesting way to still be using dogs as the gold standard, but help answer some of these questions on what is closest to their experience.
If we can find one that’s pretty close, that might then be the framework for a blood test. We’re doing blood, but it could be any substrate, like, “If you get these chemicals in this kind of a signature, let’s flag this dog for further testing.” We might not be using this as a be-all, end-all diagnosis, but people don’t have the funds nor the inclination to put every dog through scans every year. That would be incredibly expensive and a waste of resources. If you got a match on this quick test, maybe then you would want to go and do further screening. That could help people catch it earlier than they’re catching it now.
Determining When Working Dogs Reach Their Limit And Retire
How do you know when a dog has reached their limit? They need to retire, so to speak.
It’s cognitively taxing. It’s a lot of hard work. A lot of our dogs, after 5 or 6 years, stop having fun doing it. Their performance goes down, and we’re like, “Guess what? You have contributed so much. You now can go and lie on the couch for the rest of your life.”
This stuff is very complex. We have our UDC as the easiest version, but we could also do easier versions of the hemangio in the lineup, and then two empty boxes. We have those checks and balances. If a dog who used to be very proficient is not reliably giving us the answer on some of those easier questions, that’s a good tell that, for whatever reason, they’re not interested in participating anymore.
One of the dogs in our study who was in the published paper is about to turn thirteen, and she’s still incredibly up for it. She loves it and is showing us a great performance. We know that as dogs age, their olfactory system does begin to decline. They’re going to lose some of those receptors. She’s still very proficient, so it’s interesting to see. At a certain point, we might see that decline, but we can’t predict the age at which it’s going to happen because it seems to be independent.
She’s good at her job.
She loves it.
She’s aging gracefully. They’re not allowed to be on any medications, right? Don’t some medications interfere with your olfactory abilities?
Evaluation Of How Medications Interfere With Olfactory Performance
Not allowed to be on any medications is probably an overstatement. What we find is that if a dog has arthritic pain, then putting them on pain meds improves their performance. There are some drugs that do interfere with olfaction, but there’s not a lot that we have evidence that truly does impair olfaction. Certainly, all of our dogs are on monthly preventatives. That hasn’t been an issue. We’ve had dogs that have been on gabapentin and trazodone because we need to have them resting quietly if they have an injury, but it doesn’t seem to necessarily impair their olfaction.
That one we haven’t tested systematically. There are a few things that we have tested systematically. This would even be true of veterinarians who are out there prescribing medication for working dogs. Anything that we prescribe could have the potential to impact it. There is a study on metronidazole, but it’s at doses that I don’t use. Twenty-five milligrams per kilogram twice a day, which is getting real close to the neurotoxic dose, has been shown to decrease the sensitivity of the dog’s nose in some of the dogs tested.
There are lots of things out there. What we would recommend is that if we’re doing anything, even anesthesia or some surgical procedure, if that is a drug detection dog or especially a bomb detection dog. We don’t want to interfere with any of their sensitivity. Have the handlers do a test with the dog with their training aid. Make sure the dog is performing at the expected level. At this point, that’s our best bet.
If we’re doing anything—whether anesthesia or a surgical procedure—with a drug detection or a bomb detection dog, we don’t want to interfere with their sensitivity. Handlers should test the dog with its training aid to ensure it is still performing at… Share on XIf all of the dogs were trained on this Universal Detection Compound, the UDC, then they could calibrate their dog before they went out. Unfortunately, that’s not a widely available process. We have to evaluate the dog, make sure that anything we do as veterinarians doesn’t impair their performance, and have that conversation with the handlers. We know that even sometimes, the emotional stress of being in the veterinary hospital, the dog might not be on their game for 24 hours or more.
Recognition Of Famous 9/11 Search And Rescue Dogs
Dr. Otto, did you work with the famous 9/11 dog, Bretagne?
Yes, I love Bretagne. She’s Golden and the last of the 9/11 dogs in our study. She had her sweet sixteenth birthday in New York. She was a fabulous dog. One of the first puppies in our program was named after her, and it was a Golden. That Golden was a successful diabetes alert dog. That was one of the very few diabetes alert dogs that we trained.
She’s so famous in New York. We also have the famous Pale Male, the hawk. I live outside. It’s very famous. Pale Male’s former girlfriend has a partner and is back in the nest on Park Avenue. That was in the news. It’s a groupie thing here. Bretagne was also a famous New Yorker, so to speak.
I love the name.
It’s Belle Bretagne.
We need to know what pets you guys have.
We want to know.
Do you want to talk about Volk?
Volk is my foster dog. She’s a Black Lab. She’s one of the dogs in the study. She’s five years old. She’s high energy. She’s a fabulous dog. She has a strong preference for my husband, or maybe they have a strong preference for each other. They’re lovely. She is my only pet. That said, I do have a special relationship with a wild black vulture that comes and visits me. We interact. I’ve trained him to do some agility, which is silly.
At home?
I live on several acres, so he comes and visits.
I live in a little town in Ohio, which is Hinckley, Ohio, where the buzzards come back every year. I’ve yet to train one. They’re very smart.
They very much enjoy the interaction, which is interesting. I would have never known.
At OSU, when I was in Raptor Rehab, we had a couple of vultures in. They plod around, and we’re like, “Get in that bottom cage.” I enjoyed working with them.
They’re fascinating.
Dr. Wilson, do you have any pets?
I do. I’m about four years into my time in America. Back home, I have a dog with my parents who’s a Red Fox Labrador. He was a washout from a biodejection program back in the UK for environmental sensitivities. He’s a lovely pet dog. He said, “No, thank you,” to being a service dog, which was fair enough. He’s a little nervous. He’s back at home with my parents. He’s primarily my parents’ dog.
Here, I work with a lot of dogs. Outside of the postdoc, where we were also with dogs all day, I work as a certified applied animal behaviorist. I’m with client dogs throughout the week. Across all that and thinking maybe I’ll go back to the UK one day, I’ve decided no personal dogs while I’m here. I do have a cat. I’ve trained him to do the fitness program that we do here. I’ve trained him to do nose work. I walk him outside on a leash. I’ve made him into a dog. It’s easier with a cat to leave while she’s so busy during the day. Dogs are in my future.
Dr. Otto, I have a male cat that does this to me. He loves my husband and two sons. I figured that one out. It’s the smell. He runs up to me if he wants to go outside because I have a hand to open the door. Other than that, he’s like, “Get away.”
Speaking of cats, and I know we’ve got to wrap it up, remember the story a few years ago about the cat that lived in the nursing home and would go and sit in front of one of the doors of one of the residents who wasn’t going to make it. That’ll be the next project. Good luck with that one with cats.
Cats are very capable. They’re not always willing.
They don’t want to. It’s always their day off.
We have our guest question. Let’s start with Dr. Wilson. What is your pet or human peeve?
Probably still dominance theory and people who don’t always understand that when dogs are doing something, it is working for them in some way rather than them being naughty or trying to annoy you. People ascribing dogs as being stubborn or doing something to annoy the owner bugs me because that’s a strange way to look at your own animal.
Dr. Otto, what is your pet or human peeve?
My pet peeve is probably obesity in our canines because we know that it shortens their lives. Food is not love. Especially as a sports medicine veterinarian, keeping our dogs moving and keeping them physically fit is front and center.
They don’t open the refrigerator door. We’re the ones who open it.
The treat jar.
Janet, when we’re off camera, I’ve got to read to you the list of treats that they used in the study.
I use the kibble bits. It’s small, and I figured they have nutrition. You taught me that, although we do have some special treats.
I’ll make one comment about kibble. I use it all the time. We had one puppy choke on a piece of kibble. We almost lost him. It’s important in our puppies because they look up, and their chins are high, and it goes right down the wrong pipe. That puppy ended up in the ICU for almost a week. We’re so lucky to have him come back to the program and then go on to be a search and rescue dog. When we’re giving, especially to the puppies, and they’re eager and reaching out, thinking about the way we deliver those treats is important. A lot of times, we’ll drop them on the ground and let them pick them up off the ground.
I never thought of that.
They can’t help it. They’re so excited. This has been fascinating. I appreciate the work you all are doing. It’s amazing. It’s a big wow. Thank you so much for tuning in with us and being our guest. Alice, what did you want to say?
I’m in awe. Dogs helping other dogs, there’s nothing better than that. We’re fixated on dogs helping humans. They do so much for us. I’m so honored to meet you both. I was so looking forward to this.
It’s fascinating. Thank you so much. We’re going to say goodbye. Don’t forget to tune in and share. Alice, what about our website?
If you have any comments, questions, or even want to leave us a message, it is SOPPodcast.com.
Thanks for tuning in.
Important Links
- Dr. Cynthia Otto
- Dr. Clara Wilson on LinkedIn
- Penn Vet Working Dog Center
- AKCCHF May 2026 Webinar Certificate
- Trained Dogs Can Detect the Odor of Hemangiosarcoma in Canine Blood Samples
- Penn Vet Study Finds Dogs Can Detect the Scent of Deadly Canine Cancer
- PODCAST: Dr. Cindy Otto on Latest Dog Overheating Research, Working Dog Science, and More
- Penn Researchers are Training Dogs to Detect an Often Deadly Form of Canine Cancer
- Dogs Can Sniff Out Human Stress
- Dogs Can Smell Human Stress, and It Bums Them Out, Study Shows
About Dr. Cynthia Otto
Cynthia M. Otto, DVM, PhD, a tenured Professor of Working Dog Sciences and Sports Medicine at the University of Pennsylvania, School of Veterinary Medicine, is board-certified in veterinary emergency medicine/critical care and canine sports medicine/rehabilitation. As executive director and founder of the Penn Vet Working Dog Center, she oversees fitness and medical care of the program’s detection dogs, provides rehabilitation and conditioning for police and other working dogs and conducts vital research on and by detection dogs. With over 165 peer-reviewed articles, reviews, and book chapters, she is an internationally recognized expert in both emergency medicine and working dog science. She was Pennsylvania’s 2002 “Veterinarian of the Year”, received Ohio State’s Alumni Recognition Award (2006) and Distinguished Alumnus Award (2008), AVMA’s Bustad Companion Animal Veterinarian of the Year (2018), Mark Bloomberg Award (2019) and Asa Mays DVM, Excellence in Canine Health Research award (2021), International Canine Health Award, Special Award (2024), AVMA Career Achievement in Canine Research Award (2025), JSWK9 Lifetime Achievement Award (2025), and the Dr. Jack Mara Scientific Achievement Award for ACVECC: (Working Dogs) (2025).
About Dr. Clara Wilson
I am a researcher specializing in canine olfaction, cognition, and behavior, studying how dogs perceive and interpret the world. My work examines dogs’ detection of complex biological odors alongside cognitive processes such as category learning, with the dual aim of deepening our understanding of canine perception and ensuring that applications are grounded in welfare-led, evidence-based practice. Through this research, I aim to advance our scientific understanding of canine scent perception and contribute to ethical applications that benefit both dogs and people.
I completed my PhD at Queen’s University Belfast, where I investigated dogs’ ability to detect odor changes associated with human psychological stress, and I hold an MSc in Applied Animal Behavior and Welfare, during which I studied canine detection of blood glucose variation. I also hold a BSc in Psychology. My work emphasizes rigorous experimental design, interdisciplinary collaboration, and a One Health perspective linking human, animal, and environmental wellbeing.
Alongside my research, I am a Certified Applied Animal Behaviorist (CAAB), working directly with dogs and their caregivers to address behavior concerns using humane, science-based assessment and intervention.