So, good afternoon. It's a pleasure to be with you today and to show you a really innovative tool that we have developed using artificial intelligence to help detect in just one visit cats at risk of developing chronic kidney disease. Artificial intelligence is absolutely everywhere.

We may not realise this, but we use it constantly. It predicts the weather forecast you saw this morning on TV. It is in all our cell phones with auto-completing texts, and it is also determining how expensive your car insurance might be.

It's obviously a gold mine for health care, and it can now make us augmented beds, so making the invisible visible, enabling us to predict if the senior feline patient on the table without any clinical diseases may develop a renal disease in the next 12 months. But before going into further details with renal, I'd like to take a moment to discuss how digital transformation can make our life easier. It is everywhere impacting our daily life.

It is making health and veterinary care evolve pretty fast. It's obviously accelerated by the COVID crisis, and it can become a great opportunity to add value to your consultations. And at we believe we have a role to play in this transformation using digital platforms to make each and every vet a nutritional expert who can positively impact the life of their patients through health nutrition.

And to support this, we have created a range of digital applications dedicated to vets to make their nutritional recommendation easier. We know that you are overloaded by admin work, but we also know that pet owners expect a nutritional recommendation from their vets. So we have created this range of services to allow vets to define which nutritional solution is best adapted for each pet.

And this is particularly tricky, and we will see in a few minutes for chronic kidney disease patients, for which the nutritional management has to be really tailored to the stage of the disease. Our ambition is that a simple tool that can facilitate the prescription of the most adapted diet will help enrich the nutritional discussion with pet owners to reinforce your vet experts image. Also to build trust with the patroners.

We also help improve patron's commitment and compliance with a personalised prescription and follow-up. And make you gain time by simplifying your recommendation process. It is also a way to drive footfall to the practise, and there will be a clear recommendation for monitoring each patient.

But without further ado, I'd like to share with you a little video. There we go. Chronic kidney disease, or CKD is very common inside a cat.

An estimated 30% of the feline population over 10 is affected. The earlier the diagnosis of CTD, the earlier we can intervene medically and nutritionally to better maintain quality of life. Unfortunately, because cats show clinical signs very late in the process of the disease, when 75% of the nephrine are already affected, it is very difficult to identify early.

At Royal Cannon, with veterinary nephrologist of the Royal Vet College in London, we ask ourselves, how can we help practitioners to identify those cated risks of CKD as early as possible. We retrieved historical data from over 3500 cats collected before and after a diagnosis of CKD totaling more than 10,000 consultations. This data was then used with artificial intelligence tools to create a reliable predictive algorithm.

Real the tech was born. Renaldich is a digital application which can predict with 88% accuracy if a cat over 7 is at risk of developing azotemia in the next 12 months. To predict CD using Realtech, you just need to enter serum creatine, serum urea and using specific gravity into the application.

In a single visit and in less than one minute, Real adds value to the seorial check by calculating if a cat is at risk or not to develop CTD during the year. This and give you a follow-up and nutritional advice. If at risk, a catch should be followed closely with regular kidney panels with creatinine, SDMA, and urine specific rabbits.

And as a precautionary measure, we also recommend a diet moderately restricted in phosphorus and enrich in omega 3 fatty acid and antioxidant. Identifying patients at risk of sick early allows better management and maintenance of quality of life as well as honour reassurance. Developed by veterinarians for veterinarians we achieve healthier life for senior cats.

So, you all know that chronic kidney disease is a very common problem in senior cats. And as we saw in the video, it's affecting an estimated 30 to 40% of cats over than 10 years. And it's actually the second cause of death of cats.

It is a progressive and irreversible disease, and it is well known that the earlier we can intervene in CKD, the better for the cat's quality of life and longevity. But early diagnosis of CKD is quite challenging, as there is often no clinical signs until the disease has already progressed to high risk stage 2 or 3, so when more than 75% of the nephrons are already affected. There are minimal to new clinical signs in the early stages of CKD, so when we are in Iris stage 1 or early stage 2.

And the most common test used to indicate kidney function, so creatinine, does not increase. Above reference range until quite late in stage 2, so when most of the kidney function is already lost. Diagnosis is not a simple process, and even measurements such as creatinine and ADMA need to be repeated multiple times to have a proper diagnosis.

Urinalysis as well can be indicative, but unfortunately, it can be influenced by many factors and it's often not done at routine checks. So for all these reasons, in most first opinion practises, CKD will be actually diagnosed at stage 3, so when more than 2/3 of the nephrons are damaged. Internist and nephrologists, they may spot the disease earlier, but we are still missing a diagnosis that would allow the detection of Stage one patient, that would be the dream.

And the issue we have with the tests available even in the recent ones is that they are not very sensitive at very early stages of the disease. So this is why a few years ago, we have started a project to try to do better, and our discover team had the idea to use artificial intelligence to try and predict which cat would get chronic kidney disease. And thanks to a long term partnership with the Royal Vet College in London, we had access to a database of senior cats, followed over time, and which some of them eventually developed CKD.

So those data were analysed by mathematicians to create a neural network. When working with artificial intelligence, the first thing is training in this first stage. Clinical data, so from a bit more than 200 cats from the feline RBC kidney clinic and their corresponding renal health status over time.

We entered in the system to train the computer how to distinguish between healthy cats and those with CKD so from their clinical data. And the result was this initial model or algorithm. Then this algorithm was tested and validated using new sets of data.

So coming from Bandfield Pet Hospital, which is a big chain of clinics in North America. This data was used to stretch the performance of the model and ensure its robustness. This time, the clinical data is entered to test if the algorithm is correctly distinguishing which cats develop renal disease over time and which do not.

And from the 16 clinical input variables we had available for each of the consultations. The 3 most useful variables that allowed for the most predictive model. So urea, creatinine, and urines specific gravity was selected.

So along with the most useful time point for prediction, which was 12 months. So, if we recap, the renal detect algorithm was built using 20 years of clinical data from LCATs and CKDCATs. Then it analysed data from more than 3500 cats and more than 10,000 consultations.

And for each consultation, we had 16 clinical inputs, so biochemistry parameters, but also age parameters from the clinical exam, etc. And this resulted in an algorithm that is capable of predicting the risk of a feline patient over 7 years of age to develop azotemic CKD in the next 12 months from 3 bits of information, so urea, creatinine, and urine specific gravity. It may feel quite basic parameters, but what's really important to realise is that this algorithm does not look at data individually and compare it to the reference range.

It's actually a neural network that is using those data in combination. Doing complex mathematical equations on them to provide a prognosis and assess if the cat will develop CKD in the next 12 months. And the reliability of this model was assessed throughout the validation process.

And actually all results have been published in the Journal of Veterinary Internal Medicine back in 2019. The accuracy of the model to predict if a cat is going to develop as a CD in the next 12 months is actually pretty good, so 88%. That means that 88% of the time the algorithm will provide the right answer.

Then we have the negative predictive value. So it means that if the test is positive, the likelihood of the cat to develop CKD within 1 year is 92%. And Positive predictive value.

Means that if a cat is detected. At risk, then he has 53% of chance, so more than one chance out of 2, to develop CKD in the next 12 months. We also have assessed classic parameters of any test like sensitivity, which is 87%, and specificity, which is 70%.

So, in summary, after a single visit, this model will be able to tell you once you'll feel creatinine, urea and USG value. If a cat over 7 years of age is either so at risk. Of CKD in the next 12 months, which means that this cat has more than 50% chance to develop a stage 2, azoamic CKD within a year.

Or not at risk, which means that this cat has less than 10% of chance to develop a stage 2 CKD within the next 12 months. What's important to remember here is that this tool gives a prediction and not a diagnosis. With this prediction, you will be able to adjust the nutrition and better plan and tailor the frequency of your follow-up visits.

For cats not at risk of CKD, a senior diet can be recommended, and a checkup can be planned in 12 months. For cats at risk of developing CKD, we would recommend an early renal diet and organise a recheck in the next 4 to 6 months to reassess the patient. And it's frozen, but OK.

So, what is The rationale behind this nutritional recommendation for a cat, so more than 7 years old. Predicted not at risk of CKD? We would suggest a senior diet, which will address the first stage of maturity of the cat.

So in the current right can range, this is mature consult, which is supplemented with an antioxidant coffee to fight the accumulation of oxidative stress with age. It's enriched with EPA and DHA for their inflam anti-inflammatory effects. And it's supplemented with control protectors to support ageing joints.

Protein in this range is slightly higher than the corresponding adult diet to help maintain muscle mass through ageing. For cats predicted at risk of CKD within 1 year, so by the renal detector, as these cats may already be in CKD stage 1 or will likely turn stage one in the coming months. We would suggest to feed an early renal diet, so the same diet as we recommend currently for CKD stage one patients.

This diet is more directly restricted in phosphorus and protein. And for a cat which creatinine and USD were already indicative of CKD stage 2 or over. From the Irish guidelines.

If the repeated measures confirm these values, then the cat should be fed a renal diet, which is further reduced in phosphorescent protein to help control hyperphosphateemia and uremic signs. So now let's try to see how Renal detect works in practise and imagine you received today, so strik. The slide is wrong actually.

That would be a feline patient of 9 years or a senior health check. So you've done your blood work, your biochemistry and urinalysis, and all parameters are within the reference range. And here I will need to change my settings to show you this little video.

So you're on the renaltech tool, and you will enter the value of your cat. So blood creatinine. You can enter Any of the different units and the tool will do the conversion.

It is also telling you when the value you may enter is not in the current reference range. Then you can enter blood urea. And urine specific gravity.

Then you click and the algorithm is working. So this cat actually was found at risk of developing a chronic kidney disease in the next 12 months. So, the tool also gives you some links to go on the Iris International recommendation.

And you can then Apply for the best nutritional solution. So here, to better tailor the rationing and the daily aggregation, we will ask you to enter the pet profile. So the name of the pets, the sex, the neutral status, the date of birth, the current weights, and the activity level.

The algorithm will also tell you if the cat is less than 7 years old because the algorithm is only working for cats over than 7 years. Right? Then you can select the appropriate diet.

So in this case, it could be earlier enough, dry or wet. And the tool will be giving you the appropriate daily allocation. So here in this case in mix feeding, that you can at every stage, tailor.

Yourself. And you can print. The recommendation, and then you have A full recap for the pet owner.

That you can also personalise with, for instance, your clinic logo and other observations that you've made during the consultation. And if the owner wishes, you can also send. The full recap by email.

This is how the recap looks like. And from one video to another, you will be also able to track weights. And all the different parameters.

So What kind of protocols can we recommend? So you see your senior patients, so I insist on that, more than 7 years of age, and do your blood and urine tests. In case the cat is already out of the reference ranges.

You may confirm the CKD diagnosis using the Iri guidelines and. A renal diet. And then you have your classical CKD management protocol.

If nothing is apparently abnormal, then you may run the renal detection. If the cat is detected, not at risk. Sorry, not right.

OK. The cancer is detected not at risk on top of the senior diet. You may recommend a normal follow-up protocol with yearly assessments.

In case the cat is detected at risk to develop CKD, you may recommend unearly renal diet with a closer follow-up with 4 to 6 month visits. So what's the rationale behind suggesting this early renowned diets? Well, we know that classical therapeutic renal diets are widely recognised for their beneficial effects in pets with asulteic so stage 2 and late stage 2 CKD.

The question was here, how should we extrapolate that to stage one patient? If we look at the different nutritional levels we have in a classical renal diet, the main point is the low phosphorus content. We can also have additional phosphorus binders, and this is really to fight hyperphosphatemia and hyperparathyroidism.

We can also have EP and DHA supplementation that aims at reducing the inflammation and therefore the kidney lesions. Those diets are supplemented in antioxidants to fight oxidative stress and kidney lesions as well. And they usually have moderate protein content to lower uremic signs, potassium supplementation to fight hypokalemia, and alkalinizing substance to fight metabolic acidosis.

So, behind that, what we target is either A compensatory mechanism. So we Low phosphorus diet, for instance, or causing processes. So with kidney lesions, inflammation oxidative stress, or clinical signs.

So what do we need to keep for those stage one cats that are at risk? The clinic and signs are not there, but we want to address the causing process. And to some extent the compensatory mechanisms.

As in early stages of CKAD, cats need a lower level of phosphate restriction, notably to limit a risk of hypercalcemia. And the evidence behind this strategy was shown in this study, so I put the reference here that was presented at the 2019 ECVIM Congress that suggested a positive effect of diets more directly restricted in phosphorus. On calcium phosphorus homeostasis in nonaotemic senior cat.

So in those cats, you start lowering phosphorus intake, but you don't want to reach the level of restriction that is needed in a later stage patients. That's really also highlighted by Iris in a note that is explaining that for early stages patients, especially if there is a development of hypercalcemia, we want to have a moderately restricted phosphorus level. So what's the efficacy of this tool on the field?

So before the launch, this tool has been assessed on the field with a pilot study involving 10 clinics and in France and 72 cats, all from 7 years of their age, which undergone a vet consultation, a blood and urine sample. We are taking all those cats and run through the algorithm. We had also 61 owners that were asked to provide feedback, and 8 veterinarians who really participated to an extensive interview process to really give their insights and feedback on the usage of the tool.

And the results showed that vets found renal detect quite fitting to their daily routine, and they found it was a real opportunity for them to give a nutritional recommendation. In fact, 83% of cats at risk received a nutritional prescription after running the renal detect tool. On their end, pet owner felt quite engaged and reassured.

So 93% of them had a positive feelings after the visit, and regardless if their cat was found at risk or not. And 65% of cats not at risk would come back. The following year and pay for another visit, which was an interesting finding.

And another interesting insight is that 95% of owners found that the cystocentesis was worth it, even though quite a lot of them were initially stressed by this procedure. So that's quite interesting as vets often feel uncomfortable performing a cystosynthesis during a routine check, but Actually, the, the pederers thought it was worth it. So what was the results among those cats?

So among those 72 apparently healthy cats, actually, 11% of them were predicted at risk, so quite likely already in stage one, and 15% of them were already out of the reference ranges, so they could benefit from an earlier intervention. So with this tool, we believe we are creating value for the full ecosystem. So the veterinarian, the patient, obviously, and the peer.

For, for vets, Real detect brings more value to your senior routine checks for the same cost. It's also providing some earlier insights so that you can deliver a better, more targeted preventative medicine. And really value your expertise as a vet.

And it can also help strengthen your relation with pet owners because they will be entering in a monitoring process. So coming back on a yearly basis if the cat is not at risk or more frequently if the cat is found at risk. For cats, obviously, it provides an earlier care thanks to an earlier detection of cats at risk.

And therefore we can help limit the progression of the visit by the right intervention. And for the pet owners it's really providing a peace of mind in both cases actually. So either the cat is not at risk, obviously the pet owner is relieved.

But even if the cat is detected at risk, he will have a proper care, a close follow-up, and will be taken care of as early as possible to limit future damages and and preserve the quality of life of the cats. So we've reached the end of this presentation. I'm really, I really want to thank you for, for your attention.

And if you want to know more about this tool, I can invite you to review the the publication, so from the Journal of Veterinary Internal Medicine that explains the concepts and all the validation process. And you can also in this platform, review some existing webinars. So we have one with Vincent Bourges, our scientific director, and also one with Jonathan Elio from the Royal Veterinary College about the early detection of chronic kidney disease in cats.

So thanks very much and have a great afternoon.