Description

With Professor Jonathan Elliott .

Transcription

I'm going to talk for the next 40 minutes or so on early detection of chronic kidney disease in cats. So, let me just make sure that I'm able to move the slides on. So chronic kidney disease is always described as a single disease, but it's actually a heterogeneous disease syndrome, and it generally occurs in cats that are of older age, and there are lots of figures in the literature, some of which vary considerably.
But something in the region of 30% of cats over the age of 12 have some degree of reduced kidney function, and I'll talk about the terminology in a minute. I guess the challenge, therefore, is to identify the underlying disease process at that point, which is damaging the kidneys, so that that can be treated at an early stage. And to do this, we really need to detect, changes in kidney function that predict the impending azotemia where chronic kidney disease is obvious clinically.
And what I'm going to talk about is determining the risk factors first of all, so we know which cats to screen in order to identify chronic kidney disease at an early stage. And then we also need to find early markers of chronic kidney disease, which enable perhaps to intervene by taking tissue samples at a point where the histopathology is going to inform the. So at the moment, when we identify chronic kidney disease, it's at a point where really we're dealing with, a consistent pathology that is, effectively the result of some underlying disease that we don't know, what the, what that, the nature of the underlying disease is and what we're looking at really is the wreckage that's left following that damage.
So the terminology is confusing, and the Irish group have tried to unify the terminology by introducing a number of different stages. And what I've done on this slide is to overlay the stages with previous terminology. So stage one is where there is significant kidney function left, but it's declining, and declining towards the point at which it becomes clinically obvious.
So if we take a cat that has Normal kidney function, and then there is some primary disease process that causes extrinsic damage to the kidney and leads to the loss of nephrons. Then that cat clearly has early renal disease, but there's no biochemical evidence of that. Now, we're going to talk about whether we could get evidence from other sources.
If that extrinsic damage continues, then the cat may well then develop some form of renal insufficiency, which, when challenged, means that you would be able to detect, so it's not able to concentrate its urine when it's dehydrated, for example, at that point, then maladaptive responses occur to that loss of functioning nephrons. So the mass of kidney nephrons has reduced down to around about 1/3 of the original number, and that's a maladaptive response would include glomerular hypertension and hypertrophy of the remaining functioning nephrons. Extrinsic kidney damage may continue, during that, that process, reducing the nephrons further, but these maladaptive responses.
Nephrons can in and themselves lead to further kidney damage. And once the cat enters stage two, then those processes are definitely ongoing. And the next stage would be, clinical evidence of that, mild azotemia.
And the maladaptive responses continue. So below, a quarter of the kidney function left, then we're thinking that the, the cat would be in stage 3, for according to the iris, classification. And the maladaptive responses would include bone mineral disturbances that are very obvious.
In fact, our research suggests that the bone mineral disturbances should go up into the early stages, perhaps stage 1, stage 2. Where the cat really is responding to loss of functioning nephrons by activating those processes. Whether those processes are involved in causing further intrinsic damage to the kidney remains to be determined, but they certainly can be used as prognostic indicators.
When the clinical signs become really obvious in the late stage 3, stage 4, then we would describe the cat as being uremic, and many of the clinical signs are due to retention of toxins that are normally excreted by the kidneys, and this cat will be anorexic, perhaps vomiting, and, and in need of supportive measures to maintain a quality of life that is acceptable. And that's where we would see cats in stage 4. So I really, a lot of this, of the stages, stage 1 and most of stage 2 is, if you like, under the water, not detected clinically at the moment and certainly isn't detected by chemically, until, the late, later stage 3.
So what we've done at the RVC is to try and understand how we could determine the cats that are at risk of developing azotemic chronic kidney disease following screening. And we're screening cats that are greater than 9 years of age. And over the 4 year period between 2005 and 2009, we screened 343 cats, Many obviously don't necessarily return for subsequent visits, and so we lost those to follow up.
But we did have 195 cats that were followed. 39 of them developed non-renal disease, and that's one of the problems of investigating problems of older cats. They get hyperthyroidism, and they get cancer and other factors.
100. 48 of the remaining cats were followed, completed full lifestyle questionnaires, and 27 of those 148, that's 18%, developed aoemic chronic kidney disease over the period. Now, that's lower than we reported earlier in 2009, Roseanne Jepson reported on cats becoming Aote.
Within 12 months of a health screen, where we had 29 of 95 cats, and that was 30%. The main difference between these two studies, Natalie Finch published her work in 2016, is that the age of cats that were screened and in Natalie's study, the median age was 12, and Roseanne's study, the median age was 13. But what Natalie was trying to do was to identify risk factors for chronic kidney disease based on lifestyle that we could look at and combine with biochemical findings, which Roseanne focused on, and to identify cats at an early stage.
And knowledge of these risk factors, the lifestyle factors would help us to target screening of the cat population that were at highest risk. Obviously, age is one of those risks. Factors, but the lifestyle questionnaire collected data on these 148 cats and those that were significant in the univariable Cox regression analysis and for the development of azotemia where age, as we would expect, frequency stroke annual vaccination was another factor, moderate dental disease and severe dental disease.
So dental disease came out quite strongly in this study as being a risk factor for chronic kidney disease. And just to show you the multivariable analysis, and frequent stroke annual vaccination, increased the risk by about 5-fold based on cats that were only vaccinated once, in their, in their lives. Now, we're looking at a population that presents to a charity clinic, the PDSA, which don't offer annual vaccination, in addition to the Beaumont Sainsbury's Animal Hospital that do offer annual vaccination.
So our population is a little different to the standard population, attending veterinary practise where most cats would have vaccination, perhaps not on an annual basis, but, maybe on a, on a frequent basis. So perhaps this population was somewhat unique in identifying this as a, as a risk factor. The reason that it is a risk factor and it's associated is not clear, although there are some hypotheses that could be put forward.
Mild dental disease was not associated with an increased risk, but moderate and severe dental disease were, and you can see that the risk seems to be proportionate with moderate dental disease having a 13-fold increase and severe dental disease, a 35-fold. Increase in risk, but very wide, 95% confidence limits on those, which is the two figures in the second and the third column. Sorry, the 3rd and the 4th column of this table showing the uncertainty of that increased risk based on this.
She's a relatively small population study. Just putting those into survival curves, those cats that were never or occasionally vaccinated are shown in the upper curve and those that had frequent or annual vaccination shown in the lower curve. Again, these are relatively small numbers, but the curves are separating very clearly in this, in this study.
And the dental disease, it's uncommon for cats to have no dental disease, mild. These are clearly different in terms of the onset of azotemic chronic kidney disease in this study compared to those with moderate and severe dental disease. But notice the very low numbers that we have in this study where we've clearly documented the severity of the, of, of the dental disease.
And if I was to criticise this study, the standardisation of, scoring of the dental disease, Not as precise as, as we would, design a study going forward. And obviously there were multiple people who were examining these cats. So it's fairly approximate, but nevertheless, it does come out as a, strong, risk factor for development of chronic kidney disease.
And if you look at the data in other species, in the dog, for example, and in humans, the same seems to hold true. So the conclusion to the epidemiological part of this talk is that the main environmental risk factors for chronic kidney disease in the cat are severe dental disease and frequent vaccination. What we can't say is, what is it that's linking those or what's associating those with the development of chronic kidney disease, and I'm by no means, advocating that we should not vaccinate cats.
I think the health, preventive, . Benefits from using vaccination are very, very clear. Free vaccination is something somewhat of a controversial topic, and it's something that we may want to think about.
But we don't understand the linkage between between frequent vaccination and chronic kidney disease. Obviously, Preventive dental care is something that we can do to try and mitigate against the risk of of chronic kidney disease. But if you're looking at older cats, which have dental disease, then they are at higher risk and should, therefore, that's a population that we should be screening carefully for chronic kidney disease.
And, and that helps us to identify the, the, the animals that are presenting to clinical practises that we should be focusing on, our, our screening, processes on. I want to, in the next part of the talk, really talk about assessing kidney function. And I'm going to talk about assessing the number of functioning nephrons through glomerulf filtration rate and markers of glomerul filtration rate, but then go into other means of assessing kidney function that are not really telling us about filtration, but are telling us about tubular function.
And then talk about the extra renal consequences and how that of reduced kidney function and how those can be assessed. And then finally, at the end of the talk, mention a little bit about algorithms and machine learning, to combine factors, that we might be trying to combine in our assessment of a biochemical profile, for example, but to do that in a more systematic way. So, measurement of GFR is, is possible in the experimental setting.
Is to look at inulin clearance and to look at that by infusing inulin to a stable plasma concentration and then collecting urine that's produced over a time period. Now, clearly, that's not clinically applicable, but that would be the gold standard of measuring er infiltration rate. The plasma clearance methods are practical to do in clinical practise, although they do require multiple samples and multiple sampling of cats in clinical practises is problematic.
And so they're not going to be done routinely. And in order for a marker to be used for a plasma clearance method, it has to be distributed only into extracellular fluid volume. So fluid only, it has to be freely filtered, not reabsorbed and not secreted into tubular fluid.
In other words, glomme filtration is the way in which it's removed from the plasma. There mustn't be other means by which it is cleared from the body, so it can't be metabolised in other ways. And there needs to be an accurate and reproducible method for measuring the marker in plasma.
And ideally, we would like to only have to take one sample from the cat, but that's going to reduce our accuracy and confidence in the measurements. And the markers that are used do include inulin, but inulin does appear to have some biliary secretion, which makes it less than ideal. IHL, which is a radiographic contrast material, which seems to be the best method because it really does get distributed into extracellular fluid and follows the ideal marker on properties on this slide.
Creatinine, which unfortunately distributes much further than extracellular fluid. And so it's the equilibrium takes longer to achieve and you can measure exogenous creatinine clearance. And one of the problems is getting hold of a sterile form of creatinine in order to inject.
So IHL is, is, is the, the one that we would recommend if you're going to go down the route of measuring glomer filtration rate. DT DPTA is used in human medicine and, is, used if you have nuclear medicine, capabilities. So I'm not going to talk about DTPA at all.
. In at the RVC we offer a service for measuring theme infiltration rate, which is primarily used for dogs that are polyuric polydipsy, where the clinician has not managed to find a reason for that. And the kidney parameters on a biochemical profile are normal, so they're checking out, kidney function using glomera filtration rate. And we collaborate with a company called Delta Dot, that, have In the next door we see you have developed high performance capillary electrophoresis method for measuring IHEXL, and that seems to perform very well and has been validated against HPLC and which is, one of the standard methods for measuring IHEXL.
And, we're confident that it's, measuring IHEXL accurately. And so that, that can be offered as a, a method of measuring glomer filtration rate that, and, we have a protocol where you take a sample at 23, and 4 hours, and, you correct for the redistribution phase of the pharmacokinetic, profile of IEXL again, of, of, against time, plasma hexel against time, and, and that gives us a fairly accurate glome filtration rate. I think one of the things that you need to be aware of though is that glomerafiltration rate does vary from minute to minute, during the day, although we like to think that it's regulated and the auto regulation keeps glomera filtration.
Relatively stable. It will change, according to the cat's physiological, state, and it will change, according to, increases in blood pressure and changes in, in emotional state for the cat. It's not as stable as, as we would like, and we would teach the students in our, in our, in our teaching.
So, there will be variability there in the filtration rate from, from 1 minute to the next or from 1 hour to the next. And what you're measuring is that a over a 3 hour period, rather than something that's averaged over a 24 hour period when you're measuring markers of glomera filtration rate. And the two classical markers that we use would be creatinine.
And SDM now SDMA, and I'm going to talk a little bit about, about both of those. So creatinine, as we know, is a muscle breakdown product. Creatine phosphate is an energy store in muscle.
It's rate of formation, therefore, does depend on muscle mass. There is a little bit of tubular secretion of creatinine that occurs, and there is some intestinal degradation that can occur, and some foods can actually contribute to the plasma creatinine concentration and if if you're measuring creatinine following ingestion of food, and there is some problem with lack of specificity of the assays if you're using a colorometric assay to measure creatinine, there are some non-creatinine chromogens that contribute to the signal. If you use using enzymatic essays, then they are much more specific.
So those are some of the problems with creatinine in terms of its relationship to GFR . SDMA or symmetric dimethyl arginine is released on protein metabolism, and many of the proteins in the cells that are releasing methyl arginines are formed from his stones or a part of his stones. So methylated arginines are methylation of arginine in a protein is a pro translational modification.
Occurs, particularly in his stones, and more than 90% of SDMA is rely excreted. It's difficult to measure because there are isomers, and asymmetric dimethyl arginine would be one of the ice isomers that's very closely related to SDMA, but there are methods now available for measuring SDMA, which I'm sure will be discussed later on in the study. And what you need to remember with any marker of glome filtration rate is that there is an exponential relationship between glome filtration rate and the plasma concentration of that marker, because it is a first order process for its elimination and the rate constant is glome filtration rate.
So, The sensitivity of that marker will, depend on where you actually set the limit and where you determine, the lowest level of, of, of. And study, Natalie looked, Natalie Finch looked at the relationship between IHL clearance as a measure of glomera filtration rate and plasma creatinine concentration, and, and this gives this exponential picture here. The correlation coefficient was 0.6.
So there are other factors obviously affecting plasma creatinine concentration. And the dotted lines on this slide, the red one is for the lower limit of glomera filtration rate, and the purple one is for the laboratory, upper laboratory reference range that we were using for plasma creatinine concentration. The quadrant on this, on this slide has abnormal glomera filtration rate and abnormal creatinine concentration.
But those things in the lower left hand corner would actually have lower glomera filtration rate than normal, but normal plasma creatinine. So these we're not really detecting these by just measuring plasma creatinine concentration. These in this quadrant in the lower right quadrant, our normal glomera filtration rate, normal creatinine, and we get a few samples that have higher, in that we would predict for the glomerafiltration rate that we're measuring.
And, and that might be due to non-creatinine chromogens or it might be, due to high muscle mass, which is much more of a problem in the dog, compared to the cat. There is a good correlation between GFR and creatinine, which becomes linear if you look at the reciprocal of creatinine, and a very similar correlation for the cat in terms of the GFR and one over SDMA. And really the goodness of fit, we took a 10 samples where we'd measured GFR over a range of GFRs, and looked at the correlation with SDMA and GFR and creatinine and GFR, and it is very similar.
They both measuring and both surrogate markers of glomera filtration rate. What's done in human medicine is to try and make the plasma creatinine concentration more reflective of the GFR by correcting for various more for metrics, body composition, if you like, . Acting for the fact that there is more muscle mass in some people than others.
And we attempted to do this by taking lots of different measurements from different parts of the cat. We got an R squared value for the model of 0.67, which actually was not much better than the R squared value for the reciprocal of creatinine in this group of cats.
We had 52 cats in this. In the study. So taking account of body composition, measured objectively by different morphometric measurements did not improve creatinine as a marker of GFR so that we're not really able to recommend a way of correcting creatinine for the body size and composition of a cat that we can measure by doing these physical measurements in the cat.
And those are just the graphics that that explain that. So SDMA obviously is a new marker that's been around now since 2015. So perhaps we should stop calling it new if it's been here for 5 years.
And is definitely more sensitive than creatinine in diagnosing chronic kidney disease or predicting the onset of Aoemic chronic kidney disease. One of the studies that was published early on to support this was from a colony of cats, And they had retrospective data looking at 21 cats that developed chronic kidney disease, and they had 21 healthy control cats. And chronic kidney disease was diagnosed in 15 cats by persistent azotemia, and more than 3 months of duration, and then in 2 cats by calcium oxalate, nephroli forming, and in 4 cats by a greater than 30%.
Percent reduction in GFR over the period that they were following these cats. And normal healthy cats were followed just for over a six month period and had normal glomera filtration rate, so creatinine, you're in specific gravity greater than 1040, and there was three measurements to give a lower 25th percentile of the GFR . At 1.34 mils per kilogramme per minute.
So they, they really defined what was normal based on these, on these normal healthy cats for measuring GFR in their, in, in their, in their setting. . And SDMA, performed very well in terms of its specificity.
There were no false positives in this population, having a high SDMA, sorry, having a, a high creatinine and a normal SDMA. SDMA increased in a number of cases, quite a number of cases when creatinine remained normal, suggesting that it, it's more sensitive. And if you look at some of the individual cat data, you, you can see SDMA elevations occurring months before creatinine elevations occurred in these cats.
So serum SDMA increased before. Serum creatinine in 11 of the 21 cats and 4 cats, there was no serum, bank serum available to to assess whether that was the case. And in 4 cats, their diagnosis of chronic kidney disease was based on a reduction in GFR rather than a serum creatinine over, over 2.1 milligrammes per deciliter, and two cats that had oxalate stones were never aotemic.
So the median time before serum creatinine was elevated and compared to SDMA in those 11 cats was 17 months. I'm going into this study in in detail, because I think a lot of what is discussed about SDMA as an early diagnostic indicator has been based on this initial study. And one of the things clearly that is a limitation of this study is it's retrospective, but it does clearly suggest that SDMA is more sensitive than creatinine in detecting early chronic kidney disease.
The definition of reduced GFR in this population, is based on a relatively small number of cats. They were colony cats, and therefore fed all of the same diet and so on. That may not translate into the general population.
And the study isn't really designed to assess specificity as healthy cats were used, and these were not cats that were being screened for chronic kidney disease with other diseases that may elevate SDMA and that's a point that I think we need to bear in mind. So it's likely when you move into a wider population of cats that will be more false positives, where the Elevation of SDMA is not accompanied by the development of azotemia. And I think I'd recognise that and recommend monitoring of cats with SDMA between 15 and 19 mcg per deciliter, unless there are other factors that confirm chronic kidney disease, and that's certainly what Iris would be recommending as an SDMA above 14 mcg per deciliter is a reason to suspect early chronic kidney disease and to place an animal if it's persistently elevated above 14 mcg per deciliter into stage one.
But stage one is really a monitoring stage at this point in time. And the question that I think lots of practitioners would have is, so you've diagnosed stage one chronic kidney disease because there's persistent elevation in SDMA or there's elevation of creatinine within the reference interval over time, or there's persistent proteinuria, for example. How How do we treat these animals?
Well, one of the factors that seems to be occurring in that early stage chronic kidney disease, which may well have an impact on progression at a later stage, is bone mineral disturbance, and I'll come on and talk about that towards the end of the session. . And FG FGF 23 is an indicator that bone mineral disturbance is occurring.
So this is a plot of data that Hannah Sargent published last year in the Journal of Veterinary Internal Medicine, looking at those cats that had elevated SDMA, and there was a correlation between SDMA and FGF 23. And I'll go into this in a bit more detail later on. But, the proportion of cats that had elevated SDMA where FGF 23 was above 350, picograms per mil was highly significantly different compared to those cats that had normal SDMA.
And FDF 23 is a novel marker of bone mineral disturbance that I think could inform those cats that need to have phosphate restriction at this early stage, because not all cats do need phosphate restriction, at this early stage of chronic kidney disease. So moving along, our next want to talk about how we might assess that there's tubular damage occurring, and tubular dysfunction occurring despite the fact that there is perhaps a normal renal mass, and normal glomer infiltration rate, for example. And this is by looking at proteins in the urine, and protein urea as established by the amount of albumin that's being excreted, which is what we're measuring on a dipstick test, but we'll come in and talk about the limitations of that in a minute, or the tubular function where proteins are reabsorbed following filtration.
So detection of small amounts of protein in the urine does indicate that there's either defective glomer filtration membrane or regulation of glomerul filtration hemodynamics that's allowing more protein to escape across the the filtration apparatus, if you like, or there's defective proximal tubular function in rehabs. All being the small amount of protein that normally gets across this complex and filtration apparatus that consists of the podocytes, the basement membrane, and the endothelial cells with their glycocalyx. And so protein in the urine is really very important to consider and it is therefore very important to collect urine from cats to assess what's going on with their kidneys.
The problem is that urine is inherently variable and it does require multiple urine samples to be collected over a period of time to be sure that any proteinuria that you detect is persistent. So proteinuria has been shown to be an independent risk factor for progression of chronic kidney disease in human medicine and in feline medicine, and for the onset of azotemia in cats that have been screened. Albumen leaking across the glomerulus is an indicator of glomerul hypertension, which is one of the adaptive mechanisms that increases filtration at the level of the individual nephron.
And in the early 2000s, We did have a test for detecting albumin urea through a specific elizer and immuno test for feline albumin. But albuminuria and proteinuria is commonly associated with other diseases. So it was a lacks specificity.
And therefore, that was the major problem with this test. So transient albuminuria and proteinuria is common and it can indicate other systemic processes such as chronic inflammation and cancer. Just a few words about the, the standard dipstick test, which many of you who practise feline medicine will recognise is almost invariably one plus in the cat.
And one plus in the cat doesn't really tell you anything. In the human, one plus is is significant, and requires further investigation, but, cats have one plus protein neuro. The urine protein to creatinine ratio is below 0.2, is between 0.2 and 0.4 and is actually at the lower limits of 0.4 to 0.6 or can be 1 plus proteinuria on this dipstick test.
And so it doesn't really operate as a as a good screening test for cats in terms of the degree of proteinuria. We have shown that proteinuria is very common in, in cats that go on to develop azotemic chronic kidney disease, and this, these were the results of Roseanne Jepsen's study where we screened if you, if you remember, 95 cats, and we showed that having a urine protein to creatinine ratio of above 0.2, Increased your risk of developing chronic kidney disease.
If that was combined with a creatinine of above 1.6, which is still within our reference interval, and then the sensitivity of that for predicting the onset of aotemia 12 months later was 82.8%, but the specificity was was relatively low.
So having either of those gave you a high sensitivity, but a relatively low specificity. If you had to have both of those 0.2 UN protein to creatinine ratio and 1.6 creatinine above 1.6, then that increased the specificity but reduced the sensitivity.
The figure, the graphic on, on this slide is showing the status of the cats in terms of their urine protein to creatinine ratio and their urine albumin to creatinine ratio at the point of screening and their status is at the 12 month follow up time point. So you can see there's lots of overlap and on an individual cat basis, this test on its own is not, is not going to be very useful. .
So are there more proteins that we could measure in the urine that would increase the specificity and the sensitivity? Well, there are lots of proteins in the urine that could be measured. Some of the low molecular weight proteins that are normally reabsorbed by the proximal tubule, beta2 microglobulin now for one microglobulin.
And retinal binding protein could be measured, and we've looked at some of those. We've had difficulty in validating assays for, many of them, but they don't seem to add anything to albumin or total protein. There are Tubular proteins that are now being actively investigated that are released from cells that are damaged rather than a filtered and not reabsorbed, that would be Nystatins B, clustering, nag and gamma gamma GT and it it seems likely that urinary proteomics in the future would reveal a pattern of proteins that tells you a A lot more about what's going on with the tubules than just measuring individual proteins or measuring total protein in the urine.
But this is a work in progress and and and clearly longitudinal studies are needed where outcome is determined and the proteins measured at the screening are used to predict or, or, or looked at to see how well they predict that outcome. So single urinary markers can differentiate at the population level, animals that are at risk of azotemia, but at the individual level, there's such variability and the ability to get multiple samples from a cat in clinical practise makes this not a very practical approach, but I think the promise is held by more sophisticated, multiple markers in the urine where we have multiplex assays in the future that will determine that not just whether the cat is going to progress, but what type of kidney damage is is ongoing. Those will need though, longitudinal studies to help us to understand them.
So what I'm going to focus on in the last 10 minutes of this lecture is the extra renal consequences of reducing kidney function and go into a little bit more detail about bone mineral disturbances, because I think that's also a very useful thing to consider. And in the future may well help us in this early kidney disease, setting to understand how we might manage, some of those early kidney disease patients to improve, and to slow their progression to aotemic chronic kidney disease. So, as glomerafiltration rate drops, the ability to excrete the amount of phosphate that's taken in each day in the food reduces, and the adaptive physiological response to phosphate accumulation in the body is for the bones to secrete a hormone called FGF 23, which helps the body reduce the amount of phosphate that's present.
So FGF 23 does that by increasing renal phosphate excretion. Really reducing the reabsorption of filtered phosphate, and allowing that daily load of phosphate to be eliminated despite the reduction in nephron numbers. It also inhibits parathyroid hormone secretion.
So up until the year 2000, we kind of thought that parathyroid hormone was increasing phosphate excretion. But the problem is that parathyroid hormone also increases phosphate entry into the extracellular fluid by increasing bone turnover and which releases phosphate from bone stores and activating vitamin D, which brings more phosphate into the body. So FGF 23 is really helping to reduce PTH secretion or limit PTH secretion.
And reduce activation of vitamin D, so inhibiting phosphate entry into extracellular fluid from other sources. And therefore it was an important discovery that this phosphotonin existed. If we look at the distribution of FGF 23 in the plasma in a population of old cats here shown on the left, we can see that there's a right with skew to that distribution.
All of these geriatric cats had normal kidney function on, on screening, and, but we're obviously fed a very. Will diet. If we compare that to cats that were fed a standard diet that was relatively low in phosphate and high in quality, then we get a more bell shaped distribution with a reference interval that goes up to 204 grammes per mil, whereas the reference interval for the older cats went up to 700 grammes per mL.
And I mentioned that because when we did a prospective study, Natalie Finch looking at FGF 23 as a predictor of the onset of azotemia, these are the data that we found. So these are all the data at screening, but we've divided the cats. There were 60 2 cats in this study into 3 groups.
Group one. Very normal after 12 months. The kidney function remained very normal, and you can see that their distribution of FGF 23 is very similar to the cats fed the high quality diet.
Group 2 had borderline kidney dysfunction at 12 months. The poor concentrating ability, borderline creatinine and group 3 were aotemic, and you can see. That there's a highly significant difference between Group one and the other two groups, and what this suggests to us is that bone mineral disturbances are evident in early stage chronic kidney disease if we assume that group 3 had early stage chronic kidney disease and with the onset of aotemia 12 months later.
If we look retrospectively, and, and we use SDMA as an early indicator of chronic kidney disease, we divide our screen population up into those cats that had SDMA above 14, and those cats that had SDMA are 14 or below, and then look at FDF 23, we can see that those cats that had SDMA above 14 had a significantly higher FGF 23 at that screening. Their creatinine also interestingly was significantly higher, but their plasma phosphate concentration was not their plasma phosphate concentration was normal. And if we take different cutoffs for As for FGF 23, we can really see that FGF 23 above 300, there are 3 times higher, 3 times higher proportion of the population that has SDMA above 14, that also has elevated FGF 23, and Getting a cut off that indicates where we might need to intervene with a phosphate restricted diet is going to be somewhere between 300 and 500 picograms per mL, suggesting that more than half of those cats with elevated SDMA would benefit from a phosphate restricted diet, assuming that FDF 23 is elevated because of phosphate overload, which all of the indications do suggest that that would be the case.
So, maybe FGF 23 could be used to determine those cats that would benefit from phosphate restriction. It does need prospective study to investigate whether that is the case, and rolling groups of cats that are at highest risk of developing azotemia would be best, and that's where understanding the risk factors for development of aotemia, age being a major risk factor, but Other factors would help us to find a population of cats to test this hypothesis. Chronic kidney disease is influenced by multiple factors.
Therefore, and the early diagnosis and will benefit from the analysis of those multiple factors and the patterns of change over time. And looking at this in clinical practise is really difficult on And trying to figure out which of those cats are at high risk of becoming azotemia, azotemic and having clinical chronic kidney disease within a relatively short period of time. And so, What's Royal Cannon have been doing in collaboration.
To predict aotemic chronic kidney disease. And I think Vincent the she's going to talk about that model in the future, in, in, in the next talk. But just to mention that because we had lots of longitudinal data on cats that we knew developed chronic kidney disease, we supplied data to the mathematicians, at Royal Canon and, who worked in collaboration with Royal Canon, just to say, can you see anything in pattern of analyses that we're measuring that would enable you to develop an algorithm to predict those that in 12 months' time, will have established chronic kidney disease.
And interesting the things that came out were creatinine, you're in specific gravity and urea age in this instance does not does not come out, but we're limiting the screening to catheter of an older age above 7 years of age in this, in this instance. So by working with these, large data sets, from our own clinic, and then from a much larger data set from, clinics in the US, a model was developed that would enable you to predict with a degree of accuracy in an individual cat. Is it going to have chronic kidney disease 12 months later?
And clearly, you can predict, very specifically. Or very sensitively, which cats, that's likely to be, and you can change your model slightly to get the highest sensitivity and the highest specificity possible. And, those of you who are not, really, In this sensitivity and specificity will gloss over when these figures are mentioned.
But with these large data sets with the Banfield clinic and with in the US, where regular health screening is, is, is ongoing, longitudinal data were available. And machine learning was used with high performance computing, to learn the. Patterns that were seen in those cats that became aoteic and those that didn't, and to model the creatinine, the blood urea nitrogen, the age, the urine specific gravity, and to give us a specificity of 97%, so highly specific, and a sensitivity of 46%, so not very sensitive or a less specific less specificity and higher sensitivity with this modelling.
And what I wanted to do is just to explain very briefly what that specificity of 97% and sensitivity of 46% means. So it really is all dependent on the pretest prevalence. So if you're expecting 20% of your population of all have chronic kidney disease 12 months later and you screen 100 cats, 80 of those will be true negatives.
And with a specificity of 97%, only 3 will be false ne false positives in that, in that setting. This will be trues and with a 46% sensitivity, you will get 9 of those true positives being identified. And so you'll miss 11 of the cats by being false negatives with only a sensitivity of 46%.
But 86 out of those 100 cats will be correctly identified in terms of their status over 12 months. If you choose the model, which I think is probably the preferable one, where the specificity is 70% and the sensitivity is 80%. 7%, you still have the same pretest prevalence.
You screen 100 cats, 80 are negative, truly in this population, but 24 of those cats will come out as false positives. So they'll be identified as being at high risk of developing chronic kidney disease 12 months later, but they actually won't do. That.
Whereas of the 20 positive cats, 3 are false negatives. So you actually identify 17 of those 20, and therefore, in this model, 73 of the 100 cats are correctly identified, but 17 out of the 20 cats that do develop chronic kidney disease are correctly identified over the next 12 months. And if what you're doing is changing a diet, perhaps, and monitoring more closely, you're not doing anything that is really, intervening, in a very invasive way, then those 24 false positives are not put at any harm, and you're identifying the 17.
True positives and only missing three of the true positives. And I think that, probably that is the. For this predictive test, it's not a diagnostic test, it's a predictive test as Vincent will come come on and tell you about.
And, and maybe what we need to develop is how do we monitor and screen those cats more closely once we've identified them. So early diagnosis of chronic kidney disease is a goal if we can identify ways of halting progression to aotemic chronic kidney disease where clinical signs affect the quality of life. Currently, prediction is focused on the markers of glome infiltration rate, but that doesn't necessarily predict rapidly progressing disease.
And so perhaps other markers are needed, and that's where perhaps the urinary markers will come in in the future. Defining where there is active disease through the urinary proteo is that where I think the developments will occur. And cases with elevated FGF 23, urine protein to creatinine ratio, or perhaps in doxy sulphate, which I haven't talked about in this talk due to limits of time are most likely to progress and so, be beneficial.
Benefit from treatment. So if we, if we just do look at this flow diagram, if we have reduced GFR perhaps indicated by elevated SDMA or rising creatinine that you're trending within the reference interval, then, if there is, further tests done, and there's no evidence of active disease, perhaps we just still monitor those cats, . If, if we have evidence of active disease by your.
FGF 23 or indoxyl sulphate, then maybe there's active treatment that we can do. So with FGF 23, it will be phosphate restriction. To give us more information about the disease, perhaps if they have, elevated clustering or cystatin B, then that's an indication of tubular hypoxia and there may be specific treatments for that disease.
And that's where we want to get in the future, identifying the different types of progressive disease that are ongoing through our. diagnostics that we're using to screen those cats that we think are at risk that we've identified from an algorithm that are at high risk of, of developing aotemic chronic kidney disease. Can we hold that disease in the future?
And that might be through diet, it might be through drug treatment, it might be through, other, other measures, controlling blood pressure, for example, maybe another aspect. So that's what I wanted to talk to you about today. Thank you very much for your attention.
And I'm happy to take questions just to acknowledge the people who actually did the work, PhD students in our group that have graduated since, And, academic colleagues, Roseanne Jepsen, in particular, Hattie Syme, and, the clinical nurses who have worked in the group. And the funding that is predominantly, has been from the Mars group of companies, Royal Canon, Waltham centre for Pet Nutrition, and, Mars Scientific. Thank you very much.
Thank you very much. There's just one question in the chat box. I don't know if you could see it or if you would like me to read it for you.
You could read it, that would be great. Sure. Do you think the relationship with vaccination frequency might be due to the increased number of stress episodes on attending the surgery?
I, I guess that's possible. I, I, I would hope not. I would hope that, the, regular attendance on an annual basis to a veterinary practise is not something that contributes to, to disease, and that the benefits outweigh the, the harms.
Through regular examination of the, of, of the cat. I, I mean, maybe we shouldn't underestimate in, in some cats, how stressed they actually get through the transport process and, and, and, and so on and so forth, but I would hope that that isn't the case in most instances. OK, that's great.
Thank you very much.

Sponsored By

Reviews