Improving cancer care for canines and humans alike
21 Mar 2023
Prefer to listen to the interview? Head over to our podcast 'Conversations with COSMIC' to hear the full interview with Dr Wang and follow us for a new episode every month!
An expert in cancer genomics, Dr Guannan Wang has dedicated herself to enabling genomic diagnostics and precision medicine for companion animals. She has a track record of successful grant applications, publications and has developed multiple genomic tools for dogs. In her current role as Head of Clinical Curation and Genomic Reporting Lead at Vidium Animal Health, Dr. Wang is creating a best in class canine precision oncology data ecosystem to advance cancer clinical management in dogs through genomic diagnostics and biomarker discovery.
We caught up with Dr. Wang to hear how her team’s data ecosystem could go on to inform cancer research for canine and human patients alike, and precisely how COSMIC has been involved in this first-of-its-kind project.
What led you to working in this field in particular?
I'm currently responsible for building a first of its kind, best in class canine precision oncology database. I'm also responsible for the reporting of our genomic test, which is called Searchlight DNA. This is a hybrid capture based cancer gene panel test for dogs. For those who are not familiar with it, I like to think of it as a canine version of the human panel tests that are used widely these days, such as FoundationOne and MSK-IMPACT. That's what I do for my work right now, and how did I get into this field? I didn't expect this career growing up, I wanted to be a scientist, but to work with animals? I didn’t decide that until later.
I did my PhD in molecular and cellular biology with Dr. Yan Chang at Arizona State University when I first came to the States, but what really got me interested in cancer in dogs was when I did my postdoc training with Dr. David Roth at University of Pennsylvania. So Dr. Roth built one of the first genomic diagnostic centres in the United States, called the Centre for Personalised Diagnostics at the time. Not only is Dr. Ross a pioneer in human precision medicine, he is also a proud father of three dogs! This was back in 2013 to 2015, when precision medicine was first taking shape and changing the human oncology paradigm. I also grew up with dogs in my house, so both being hardcore dog lovers, we were naturally hoping to see what we could do for dogs. As we know, cancer in dogs is very, very common, and a leading cause of death. The rest just happened, and with the help of our collaborators, we published some of the first genomic studies in canine hemangiosarcoma and found actionable mutations.
There's a few moments in your career and your life that you remember, and I think that's one of those. When I first saw those PIK3CA hotspot mutations, recurrent in dogs in hemangiosarcoma patients. It was just unbelievable!
I saw the potential, and the need, from both the dog side as well as the human side. Then, I decided to dedicate myself to applying genomics to better the lives of companion animals.
This is also part of the reason why I joined Vidium last year. I got to work with some of the top minds in the veterinary community, as well as canine genomic fields, and now we're world experts! We have the shared goal of enabling precision medicine in veterinary oncology, and giving the dogs we love the fullest life possible.
For those who might not be so aware, would you be able to tell us a little bit more about canine cancer, such as its frequency, risk factors etc?
Not until I really got into this field did I get to know how common it is. Take the current status in the States, for example. We have around 90 million pet dogs in the US, and 6 million of them are diagnosed with cancer each year. That's a huge number, more than 6%, and the most common ones are lymphoma, mast cell tumour, hemangiosarcoma, osteosarcoma, and melanoma, just to name a few. Some of these are more common than their human counterparts, for example, hemangiosarcoma is a very, very common cancer in dogs and very deadly, but the human equivalent (angiosarcoma) is a very rare cancer. One of the reasons for studying hemangiosarcoma was that we could potentially use that information to help drive research and clinical advancement for angiosarcoma patients as well.
What are the current treatments that are in common practice for dogs? I mean, is it the same as we see in humans, such as radiotherapy, surgery and chemotherapy? Or does it differ?
Both the clinical presentation and clinical management is very similar between humans and dogs. We have radiation therapy, surgery, chemotherapy, we see a few targeted therapy options and immunotherapy is also in the works. In terms of these examples though, they’re approached differently as I learned from my veterinary oncology colleague, Dr. Esther Chang. For dog patients, you don't use the highest dose, so you use the maximum tolerated dose. This is to allow them to experience the efficacy without making them too sick. The doses used in humans are much higher to ensure efficacy at the expense of a severe side effect. For human patients, you can explain the risks and adversities to them, and have them choose whether they want to move forward or not. This isn’t the case for canine patients, it is hard to explain these complexities to them, and their owners rarely want to see their dogs suffer from the side effects of chemo from radiation. So, this is why the maximal tolerated dose is preferred for canine patients.
Are there clinical trials for dogs, and how do you manage the ethics around treatment?
Absolutely. There are many, many, many clinical trials. One of the features of our genomic reports and our knowledge base (obviously, besides the mutation biomarker associations) is that we scan the entire clinical trial space to include any and all ongoing clinical trials that are turning to therapeutics for canine patients. We currently have 173 active clinical trials that offer therapeutics to canine patients. So definitely, there are a lot of clinical trials.
What is the current status of personalised therapies for dogs? Is genetic sequencing frequent in canine patients or is it only in special cases?
There's one FDA approved personalised cancer therapy for dogs which is a tyrosine kinase inhibitor which treats canine patients with mast cell tumours. Additionally, there are several targeted therapy clinical trials that are ongoing, and I'm honestly surprised that there aren’t many more. We actually have several that are in the works that are about to be released. So, I'm glad to see that there’s definitely progress there.
Genomic tests definitely aren’t standard, but with our own genomic tests, if we find druggable targets we offer that information. As some drugs are available, the veterinarian is often able to reach for that. Genomic sequencing is more and more popular, so I think we're still at the early adoption period, but we definitely see a higher adoption each year.
Can you tell us a bit more about the database of mutations in canine cancers your team is building?
We are building what we call a ‘Vidium Insight Data Ecosystem’ because it's more than just one database. It's a canine precision oncology database. It has several components, and we're expanding them.
It comprises of a knowledge database, where we curate mutation data, and the mutation level biomarker associations from peer reviewed literature. This is much like what you do at COSMIC. And we also have our internal variant database, from all our genomics sequence cases. We have accumulated quite a lot of mutation data that way as well. In addition, we have a clinical follow up/ outcome database, and this is where we collect real world clinical data, like treatment rates, response, prognosis, outcome, and how they respond to their targeted therapeutic drugs in the context of high quality mutation data. This will enable us to identify a lot more biomarker associations than what's published in the literature. We're very excited about this database. We’ve got to a time where we are building a bonafide data structure that will host all this data and have capabilities to pull data from outside data sources.
The word ‘ecosystem’ is such a great descriptor of this database, it really draws attention to how broad the spectrum of data is and how it can still grow.
I think we are only just scratching the surface of canine cancer genomics, so there's so much more to come. One tiny bit of difference from COSMIC is that if a germline mutation has for example, a cancer predisposition, we also include that. So we are not only curating somatic mutations, we also curate high level germline mutations.
This is where we'd like to say that canine data is increasingly abundant, but currently we very heavily rely on inferences from the human data. That's why we love COSMIC and we use COSMIC data. For our canine precision oncology database, we have one arm scanning the entire canine literature and then we bring in the high level human mutation evidence.
You said broadly how you use COSMIC and it would be fascinating to know more about how you're using COSMIC. But also how much similarity is there between dogs and humans? Can they be compared?
More similar than you’d ever think if you hadn’t got to see it actually!
We recently submitted a paper where we were able to convert the mutations found in the 700 COSMIC Cancer Gene Census genes into their canine homologues. We saw over 99% similarity, which is just amazing! This tells you how conserved the mutations are across species. Not only do we see the sequence conservation between the two species, we also see that a lot of the mutations found in human cancers are also found in their canine parallel cancers. For example, you find a PIK3CA mutation in human mammary cancer, and a lot of canine mammary cancer is also driven by PIK3CA mutations. We have established that canine cancers and human cancers are similar in their aetiology and their presentation, but now at the genomic level we are really getting to know that they're very similar, having the same drivers and the same altered pathways etc.
When you were first going about this research were there any concerns that people may misinterpret it as ‘animal testing’, and how did you navigate this?
The main differences between dogs as a model system versus other other animals as a model system, cancer in dogs happens spontaneously, it's naturally occurring. This is very different from human engineered mouse models, right? So there's nothing experimental in this, the disease is real and the pain is real. There's a dire need in the vet community.
Other reasons that dogs could be a better model, not only that the cancer develops spontaneously, but the immune system is also intact. It's different from the mice, right? It has the whole functional immune system. Also, humans and dogs share the same environment, sometimes the same lifestyles as well. We live in the same house, we may sleep in the same bed, and we sometimes share food. Growing up, my dog did eat our same food! They are exposed to the same allergens, to the same food components, to the same environmental chemicals, so this makes them great stand-in models for humans in so many ways. But also, the exposure of these similar things to their immune system has also exposed their immune system to function much more similarly to diseases like cancer than mouse models. I can’t say enough of the benefits of using dogs as a model. There's nothing experimental to this, there is the need for the veterinary community. We don't have enough good drugs to treat these poor patients, so there's benefits to both sides.
You could say it's far more comparable to a human clinical trial than it is to animal testing.
Absolutely! If you look at the very long and super costly drug development cycle for human drugs. There is pre-discovery, discovery, preclinical, and animal testing. But we have this perfect patient population with these actionable mutations, that look very much like drivers that could really respond to these drugs. Importantly, the dogs need the drugs. Oftentimes healthy beagles are used for the PK/PD testing, the healthy beagles are animal testing, that's experimenting. With our research, you do not have to subject the perfectly healthy beagle dogs to that testing, because we have diseased dogs that actually need the drugs themselves.
So it sounds like the same ethics as phase one clinical trials in humans being carried out with people who are very sick. They won't test it on healthy patients or people who don't have cancer, because it's either a last resort or because they don't exactly know what's going to happen, but it is a tiny bit of hope. Of course with both, it's about improving things for the future and trying to make that difference.
What we want to push is that here we have more and more compelling genomic data, and well defined patient populations that can be used. So we need that bridge to make these canine patients visible in drug development.
Are there any plans to extend this research to look at the differences in mutations across breeds with aims of creating even more personalised treatments?
Absolutely, yeah! So that's another benefit. I have so many benefits of using dogs as a model. Another one, as you can imagine due to breeding structures, the dog genomes/ genetic genomics are actually much simpler than your patient population. So we don’t really have that somatic and germline interplay. If you have simpler genomes, it can make your interpretation a lot easier, compared to some of these multifactorial, complex genomic scenarios.
There's some indication, from our work, that some breeds do tend to have different somatic profiles with the same cancer types. I'm hoping that as we have more genomic data, as we sequence more tumours, we can potentially see an even more striking trend of what breeds tend to have what kind of mutations. It's definitely something that we are thinking about especially since cancer in dogs has a strong breed predisposition. Some breeds tend to develop certain types of cancer, for example, a German Shepherd has a high likelihood to develop a hemangiosarcoma, a Bernese Mountain Dog, on the other hand, has a high probability to develop histiocytic sarcoma. Both are very deadly cancers. The Golden Retriever, poor Golden Retrievers, they’re so beautiful, but they tend to develop quite a few cancers, like lymphoma or hemangiosarcoma. So there’s definitely a strong breed predisposition.
Is there any understanding as to why that happens between different breeds? Or is that still being investigated?
It's still being investigated. There are hints in some breeds that some deleterious alleles/ genes have to be enriched, to potentially cause them to develop that type of cancer. So we're still trying to understand that.
So interesting how it varies between breeds. COSMIC is full of pet lovers, I have a house rabbit and a cat myself. It makes me wonder, how could this be applied to other species as well?
Great question! So dogs develop cancer a lot, and cats are another companion animal where cancer is a big issue. Cats tend to develop lymphoma as well as some mast cell tumours. So the spectrum is similar to dogs, but not as not quite as prevalent as in dogs. Horses also developed cancer like melanoma, so those are the other animals that we can potentially look into. Also, the feline genome is actually very well annotated so, when we decide it's mature enough, I think we are in good shape at the genome level.
In your view, if everything goes to plan, and everything's successful, where do you see this going in 10 years? What do you hope you've achieved?
I'd like to see genomic diagnostics become part of routine clinical workup. Just like what is happening in the human oncology field right now. I'm dying to see more drugs, and better drugs made available to our companion animals. We are still using a treatment regimen from 15-20 years ago, and the prognosis is the same all these years later, so I'd love to see a lot more advancement there and this can be largely driven by the genomics as well. I’d also like to see a cross species database and more of what we’re seeing right now where we are leveraging the well established annotation from human databases. Equally, I’d like to see the reverse! We can leverage the rich canine genomic and clinical data to help human patients. I think it's a truly mutually beneficial endeavour and relationship. I don't know if it can happen in 10 years but eventually, we want to see that diseases/cancer are no longer an issue. It's a dream, but I'm very hopeful and optimistic.
You mentioned as part of those goals, trying to build a hybrid database. One of the last questions we have for you is do you see COSMIC as being a part of the production of that database?
Absolutely, that's a no brainer! So we have big plans to utilise COSMIC, I'd love to collaborate with you on that. We are forging ahead with our next phase of converting all mutations, several million of them, into the canine genome. We'd love to leverage the actionability project as well, I think COSMIC did a fantastic job there so we’d love to tap into that. That data will enable us to do a lot more. For example, whenever we find a mutation in a patient, but we don't know what this mutation is doing in the tumour, we can cross reference with COSMIC data to see if this mutation has been found in any human equivalent or not equivalent human cancers. That gives us much more confidence that said mutation could be a driver for the tumour as well. This works the other way around too - our data can potentially help COSMIC. We could say “okay, next time you curate an equivalent of this mutation, we have a high confidence that this mutation could be pathogenic, could be a driver etc”. So that's where we can really help each other out! I'd love to see that happen.
This is such an exciting area of research that’ll benefit so many different parties. We have so many animal lovers in the COSMIC team and due to the pandemic we all got to meet everyone's pets, it's quite a nice shift. So before we let you go, I wanted to ask what's the best way to keep up to date with this work?
Reach out to us! We are constantly publishing abstracts, so you can definitely keep up with us from our publications. We just recently submitted an abstract to AACR 2023, so we'd love to report some more progress there as well. We're really proud of our scientific footprint, and we are constantly updating our publications.
Before we go I just like to give a shout out to my team. My team members who are involved in this, especially Dr. Will Hendricks, Dr. Sakthikumar and to many, many others I'd like to thank for their continued support, dedication, as well as smart thinking all along the way. I could not have done any of this without them.