Slide 1

Good morning, many thanks Fredrick. Well, I will try to give an overview of the COSMOS. I will try to concentrate on three issues.

Slide 2

For those of you who don’t know exactly what is COSMOS, then the most important part, the European scenario in stage 5 CKD and with some demographic patterns of current nephrological clinical practice and just a few flashes about preliminary view of some COSMOS outcomes.

Slide 3

As you know, the COSMOS is a multi-centre open cohort and observational study. It’s prospective, 3 years follow up, it’s European focused, 21 countries are participating in this study. The size of the sample is finished, it’s 4.500 patients from 227 centres spread geographically, medium-large hospitals and satellite units. The sites and patients, 20 patients per centre, had been randomly selected and this is a very important point that we randomised all European dialysis units and we selected the patients afterwards.

Slide 4

The main objectives of COSMOS, there are many; to investigate K/DOQI target achievements in a representative sample of European hemodialysis patients and also to estimate the association between these K/DOQI target achievements with two outcomes which are mortality and hospitalisation.

Slide 5

Here we have the countries participating and here we have the list with the sites and you can see the number of sites is proportional to the hemodialysis patients of each country, bigger countries give more patients to the study than smaller countries.

Slide 6

I will show the data. Some of them will be green, these are the global figures, I also use the red and blue, the red in some graphs to show the less than 1 year hemodialysis patients (new patients) and the blue for more than one year on hemodialysis. Then I’ll show you some other results dividing Europe in regions, artificially we reproduced two regions; western countries will be in grey and eastern countries in red. Then non-Mediterranean and Mediterranean countries; non-Mediterranean in blue and Mediterranean in yellow just to play a bit with some comparisons in the study.

Slide 7

We finished the recruitment at the end of 2007, as I said there are 4500 patients in COSMOS enrolled.

Slide 8

And this is the distribution according to the countries. You can see here Germany, France, Italy, Spain, the bigger countries contribute with more patients but all countries are represented in the study and the study doesn’t want to make a comparison between countries.

Slide 9

The idea, the aim is to do a whole European lookout of the scenario.

Slide 10

Now, you’ll see some of the results. First of all the baseline vintage years on hemodialysis. The sample we got is less than 20% less than 1 year and you can see here 29% more than 5 years and in the middle from 1-5 years on dialysis. That means that at the beginning of the study we got almost a bit more than 1/3 of new patients at one year and 2 years and almost 2/3 old patients with more than 2 years on dialysis. As in the study the 20 patients per centre are always 20 that means that each patient leaving the study is replaced by a new incident patient randomly selected. We calculated that at the end of the study in 3 years we’ll have almost a picture of 50% each side. I mean 50% new patients less than 2 years and maybe 50% of patients more than 2 years.

Slide 11

The primary aetiology of chronic kidney disease is here. As you can see, we observed like everybody diabetes is 20% (I’m talking about aetiology not diabetic patients on hemodialysis, the aetiology of CKD). Second nephroangiosclerosis, then glomerulonephritis and so on.

Slide 12

But if you compare new patients (again in red) and old patients in blue you can see a great difference, almost a 9% difference in diabetes as aetiology in new patients less than 1 year. It is an important difference. We’ll come back to this afterwards.

Slide 13

The age in COSMOS. You can see 90% of our population is older than 45 and almost 60% of the population is older than 65. That means that the mean age of the COSMOS now is almost 65 years old. There are no clear differences in vintage with less than 1 year and more than 1 year. As you can see, both are close to 65% but what is clear is that most younger patients are in the former eastern countries and you can see that the eastern countries had most people less than 45 and between 45 and 64.

Slide 14

Now, the diabetics on hemodialysis (not as aetiology, diabetic patients on dialysis). As you can see, here “YES” means 30% of diabetics and 70% of non-diabetics and if you compare with diabetes as aetiology there is a jump from 20% as aetiology to almost 29.5%. It is a 9% jump between the aetiology and the diabetes presence on haemodialysis. There are differences in diabetes and you can see eastern countries have a lower percentage of diabetes than western countries that’s 7% and a similar picture when you compare Mediterranean and non-Mediterranean countries: still a 7% difference. The Mediterranean countries lower than the northern countries (non-Mediterranean countries).

Slide 15

Then we go to compare now diabetes on hemodialysis and vintage. This does not mean diabetes as aetiology but diabetic patients on dialysis, and the difference is even bigger. This is more than 11% increment in new patients. That means that almost 40% of all the patients with less than one year on dialysis has diabetes in this sample compared with those who have been on dialysis more than 1 year in blue.

Slide 16

We know that the inadequate control of mineral metabolism in CKD is very important and one of the main objectives of COSMOS is to look for this.

Slide 17

Then I will show you just a few phosphorous data and you can see here. This is less than 3.5; between 3.5 and 5.5 and more than 5.5. Then you can see more than 40% of patients have a serum phosphorous higher than 5.5 and almost half of the sample in the range, in the desirable range and you can see here that really there are some differences. In this case non-Mediterranean countries have 0.64 serum phosphorous higher compared with the Mediterranean countries and it’s significant. Exactly the same when we compare eastern countries with western countries, it’s a 0.5 increase in eastern countries compared with western countries.

Slide 18

When you look at PTH, the overall PTH at baseline is 320 and there are differences between Mediterranean and non-Mediterranean countries. This means that also Non-Mediterranean countries have a lower PTH level than the Mediterranean countries.
Well, the causes could be multiple. You can make an hypothesis and you can say can we attribute part of the differences to a technical scene for example different strategies for blood withdrawal? Or different patterns of medical practice that is more likely?

Slide 19

Well, you see there are two ways to take blood for the analyses; mid week or post weekend and when we analyse how the COSMOS sample is, in 55% of the centres they took blood post weekend and 45% mid week that is a longer period as post weekend. Does this difference in strategy influence the results because it’s one day more in post weekend? Well the answer is yes. How could it make a difference? You see mid week and post weekend there is a higher phosphorous post weekend as you can expect and in this case a lower PTH post weekend something which you did not expect but it is lower at post weekend.

Slide 20

Well this can be due to this but also to the use of different drugs and as you can see, there is a difference in the use of active vitamin D metabolites between Mediterranean countries and non-Mediterranean countries and as you remember non-Mediterranean countries have a lower PTH but also a higher consumption of vitamin D metabolites, 6% more which is statistically significant. Also in the non-Mediterranean countries they performed parathyroidectomy with lower levels of PTH what is here between 700-1000 and this is more than 1000 in the Mediterranean countries we go for parathyroidectomy later on as a general policy even though there are not big differences.

Slide 21

Well, when we don’t control well the calcium and phosphate metabolism, we are aware that we could increase or we could modify vascular calcifications, fractures and mortality.

Slide 22

Then even though this is baseline analysis we will be looking for this in the future and then we know that at least the methods to measure, to diagnose vascular calcification are not homogenous around Europe. There’s a poster here and also in the session showing the great differences throughout Europe about this particular issue but in any case 90% of the centres used mainly x-ray and echography to diagnose vascular calcification and a small minority used a more sophisticated techniques.

Slide 23

If you concentrate on x-rays we asked in the COSMOS one question and that  was to know about the homogeneity of the sample. How often is a routine bone x-ray profile carried out in your unit? You see here 2/3 of the COSMOS sample do not routinely perform x-rays and then you cannot diagnose vascular calcification in the same way if you do it in any frequency in your unit.
Then if you divide the routine x-ray evaluation in yes (those who perform independently every year or second year) and those who do not perform routinely, you can see that in blue those who perform routinely x-ray evaluation diagnosed more cardiovascular calcification because they got of course more data related to this issue. But in some way they express the clear lack of homogeneity in this procedure throughout Europe.

Slide 24

What about bone fracture? This is an even more complex issue and the percentage of bone fracture in patients within target was very low and outside target even lower PTH, lower calcium. Outside in this case means either lower or higher than the target of K/DOQI and there’s a higher percentage of fractures. This is just looking at PTH. If you add something like looking at PTH plus calcium and phosphate, you’ve got a similar figure that is within target less fractures, the study is recording fractures compared to outside target that is either in the lower and upper limits of the target.

Slide 25

Now, the idea is to have a quick look through the features, through the outcomes and then really to have 3 years follow up. The study of course has not reached that point.

Slide 26

Now (in May 2008) 3.600 patients that is 80% of the sample has 24 months, 2 years of follow up but we have then a first analysis of one year and a half (18 months) of follow up in almost 56% of the samples (2500 patients)  and I will show you just some preliminary results.

Slide 27

This is the K/DOQI global baseline achievement in the whole sample. The baseline was finished in 2007. You can see here the figures and you can ask yourself, are there any differences between the new patients and the old patients that are red and blue? You can see here there are no differences. They are really quite similar (patients with less than one year on hemo and patients with more than 1 year on hemo) and there is no difference.

Slide 28

When you compare baseline with the 18 months of follow-up there is a very small improvement in calcium and phosphorous, 76% compared with almost 69% achieved the K/DOQI targets and in calcium from 50-54. This is a small jump but is still a preliminary and half of the sample and half of the time.

Slide 29

What about in 18 months about death? All cause 22.8 that is 15% per year and half of them are cardiovascular. This is the total number of deaths and this is cardiovascular death, it is almost what we expected.

Slide 30

If we try to analyse these deaths and put into relationship with the metabolic parameters in 18 months they were related only with high calcium, low phosphorous, high calcium phosphate product and low PTH. This is mortality all cause and cardiovascular.

Slide 31

When we look at hospitalisations well again high calcium is significant and here high phosphorous appears and then calcium and phosphorous appear again in both and again for the first time here high PTH appears as a cause of an increase in hospitalisation.

Slide 32

Then as a summary, the mean age at baseline complete baselines of COSMOS is 65 years, younger in eastern countries. 1/3 of patients have less than 2 years on hemodialysis and 2/3 more than 2 years on hemodialysis. Diabetes as a cause of CKD is 20% but however, percentage of diabetic patients on HD is almost 30% and almost 40% in the new hemodialysis patients that is one year on HD. Centres measuring biochemical parameters post weekend (55%) show lower PTH and higher serum phosphorous.
Non-Mediterranean (that means northern countries) showed lower sPTH levels because they used more vitamin D, they performed parathyroidectomy with lower serum PTH levels compared with southern countries.

Slide 33

There were no clear changes in K/DOQI achievements even when you compared new patients with patients with more than 1 year and also when you compared baseline with the 18 months follow up there are just small changes and finally, after 18 months of follow up with half of the sample of the study the mortality rate was 22% in 18 months, 15% per year. Cardiovascular mortality was 10.8, 7.2% per year and after these 18 months of follow up mortality was associated with low PTH, low phosphorous and high calcium and calcium phosphate product.

Slide 34

I want to thank Jose Luis Fernandez who is here, Dylan Rosser and the COSMOS team for the work done (we have just finished the collection of 2007 baseline sample just at the beginning of 2008 and the first analyses are already done).

Slide 35

And also all the national coordinators of COSMOS and of course the sponsors of the study. Thank you very much.