CME Slides Forum - P.A. Kalra

A joint Congress by ERA-EDTA and ISN

ASTRAL STUDY: RESULTS AND IMPLICATIONS

Philip A. Kalra, Salford, UK

Chair: Alain Meyrier, Paris, France

Marius Miglinas, Vilnius, Lithuania

kalra

Dr Philip A. Kalra
Department of Renal Medicine
Salford Royal Hospital
Salford, United Kingdom

Ok so it’s a pleasure to be here and thank you Mr Chairman, Ladies and Gentlemen. I’m going to talk about the ASTRAL study. I’m a nephrologist from Manchester and also the lead nephrologist in the ASTRAL study. We actually presented the initial data of ASTRAL at the ERA meeting in Stockholm last year. So I’m going to sort of follow that up with more ---- data and also I’m going to discuss some implications of our data and then of course you’re going to hear from Jaap Beutler about the STAR study which has just recently been published.

So first of all to emphasise we’re talking about atherosclerotic renovascular disease and not fibromuscular disease which we all recognise are two totally different entities. Here as you can see, we have much damage within even the apparently normal kidney contralateral to a stenosis, and often we have bilateral disease as well. Whereas with FMD we expect to have a very well preserved parenchyma and I think that is very, very important in terms of looking at outcome.

So why did we need a trial? Why did we need more than one trial? Because of course, you know the STAR, ASTRAL and of course there’s CORAL ongoing, well we know from US MEDICARE data over the decade 1992 -2002 there was something like a four-fold increase in renovascular diagnoses in the US.

Another point was with the advent of relatively straightforward endovascular techniques like angioplasty and stenting there was also quite an increase in the use of revascularisation techniques really peaking towards the end of the last decade as you can see. Recent data suggests that at least 30.000 American patients are undergoing some form of renal revascularisation therapy. It’s a lot of patients there are a lot of people are still getting this treatment. So we really need to know does it actually benefit them? If it doesn’t benefit all of them, how can we try and decide which patients are the ones that benefit and which patients definitely do not benefit as Professor Wilcox has implied.

One thing we do know though looking at trends in the US Medicare data is the massive impact that endovascular techniques have had. So if you look back in 1992 a third of the revascularizations were surgical, whereas by 2002 that figure what just down to about 2%, the remainder being stenting procedures. So it’s a big impact.

So that said what are the ways in which revascularization might be considered beneficial for patients? What is about their clinical presentations with renovascular disease that we might wish to improve? These are straightforward but there are many as you can see. So a patient presenting with ‘flash’ pulmonary oedema we would consider it. Those who present with acute renal failure, acute kidney injury and serious renovascular disease. We’ve known for many years that there’s been an interest in trying to maintain renal functional mass by preserving that, by trying to prevent occlusion of a kidney. There were two studies as you know one from the 1980s and one from the early 1990s showing that a kidney was lost if there was a greater than 70% stenosis maybe 10% of patients developed RAO each year.
What about the patients with rapidly deteriorating renal function? Again Professor Wilcox has alluded to those. Severe hypertension, cardiac congestion, people with congestive cardiac failure and RAS. Perhaps a third of elderly patients with heart failure may have some degree of renal artery stenosis. Then of course, the patients who do manifest a drop in renal function with ACE inhibitors. These are all possible indications.

So what is the evidence? Well, we now have data from 5 randomised controlled trials but there were these three sort of older studies relatively small 210 patients between the three of them and published a decade ago. Then of course we’ve got these studies and the CORAL study largely based in the US but now expanding across the world which is planned to finish recruitment at the end of this year.

From this evidence what do we know? Well, renal function looking at the three smaller previous RCTs because obviously I’m going to tell you about ASTRAL now they were not powered to show any improvement in renal function and certainly didn’t. We got a sense from those three trials that systolic blood pressure control might be improved with angioplasty. It was angioplasty only in those days back in 1998, 1999, so in one or two of the studies there was a modest improvement in the angioplasty arm in systolic blood pressure manifest as a reduction in tablet dosage.

This is what really underpinned the development of the ASTRAL study where you can see what it stands for.

So the main question that we asked in ASTRAL was what is the effect of renal revascularisation upon the primary endpoint of renal functional outcome? We powered the study to look for a 20% difference between the two arms, one was revascularization and medical therapy the other was medical therapy alone and at 80% power we reckoned that we needed 750 patients. Now those of you who have been involved in big trials like me will always query the fact that we base massive trials upon such tiny pilot data to underpin the power calculation. There was very, very little information out there about what rate of change of renal outcome we might expect.
Secondary endpoints: survival, macrovascular events and blood pressure control and we do have some cardiac substudies where we’ve looked at MR of the heart pre and post in the two groups as well.

So here’s the trial schema. If you had a patient with unilateral bilateral significant renovascular disease where revascularization would normally be considered and that’s a very important point because I think there’s been a little bit of confusion about the entry criteria. If revascularisation would normally be considered in that patient and there was no absolute indication to definitely revascularize them or definitely not to, then the patient was eligible. Then they were randomised on a one to one basis to angioplasty and stenting with medical therapy or to medical therapy.

We opened recruitment in 2000 and we finished recruitment 7 years later when we had 806 patients from 58 centres, 54 UK, 4 from Australasia. You can see that it was a 1:1 randomisation between the two arms. This data I’m showing you is the 33 months mean follow up data.

So baseline characteristics typical of a group of renovascular patients, 70 years old, 2/3 male, 3/4 either current or ex smokers, 30% diabetic, lots of comorbid macrovascular disease as you can see heart disease, coronary artery disease in half.

Average creatinine 180µ/L, GFR around 40 ml/min, blood pressure at randomisation 150/75 mmHg.

70% of patients had an eGFR less than 50 which was important in a study looking at renal functional outcome primarily and 3/4 of patients had RAS greater than 70%. I’ll say a little more about that as we go through.

What about concomitant medication at baseline? Well, as you would expect almost 100% of patients had antihypertensive therapy. Note that relatively low ACE inhibitor, ARB usage but this of course started in 2000 opposed to I think if we started the trial now we probably would see a bigger picture of more patients on ACEs or ARBs but on average 2.8 antihypertensive drugs per patient.

What about concomitant medications? Anti-platelet therapy in ¾, largely aspirin, cholesterol lowering treatment in 80%. 96% statins. I think again now we would probably see even more patients on these treatments.

Now here we are on to the results. What about the compliance with the randomised treatment? You’ll see that in the revascularized group of patients 80% of them actually truly underwent a successful procedure so this was 323 patients.

When you look at the 3 previous RCTs in the biggest one of those the VAN JAARSVELD study less, then I think about 50 patients underwent revascularisation. You’ll notice that it was not attempted in almost 60 patients, so in about 20% of the patients. The reason for that was that we screened for renal artery sclerosis on the basis of CTA and MRA and sometimes there was slight overestimation of the degree of stenosis, so that when the patient turned up on the radiologist’s table for the procedure the disease was found to be insignificant. There was a low cross over rate, only 4.4% of patients later in the trial went from the medical arm over to the revascularisation arm.

So here’s the main outcome measure. This is repeated measures analysis over time looking at serum creatinine and you can see down the bottom figure this is the treatment difference between the two groups and there is none by repeated measures. We can see that the power of such a big study is large early on but not surprisingly, as we get later out to 4 and 5 years over here, then obviously numbers are dropping off but we will have in 4 or 5 years time a lot more follow up at these points. But certainly in terms of the primary endpoint there is no difference between the two treatment arms.

Systolic blood pressure, as you can see in both arms there was a 5 ml drop in blood pressure by 12 months and then there was a steady improvement in blood pressure over the course of the trial so that by 5 years the average drop was about 12 ml, but again look at the difference between the 2 groups nothing.

and what about renal events? This was a composite of renal failure dialysis transplantation, no one had a transplant, nephrectomy or death. 22% events by 5 years but no difference there’s the p value between the 2 groups.

Cardiovascular events no difference 50% of the patients in both groups had these macrovascular events by 5 years.

Mortality, if you asked your child to draw two overlapping lines they would be hard pressed to do a better job than this and you can see that by 5 years 40% of patients had died in both groups. So we’re looking at a mortality rate of about 8% per year. So I’d like you to just bear that one in mind when I come to my conclusions.

What about change in therapies? By one year surprisingly there had only been a very minimal increase in use of ACE inhibitors and especially so in the revascularized group it only went up from 47 to 48% which is surprising. A slight increase in anti-platelet therapies and in cholesterol lowering so almost going towards 100%.

We did have some pre-specified subgroups for analysis I’m not going to tell you much about these now, we haven’t fully analysed these but in any big trial like this even in such a large trial these subgroups have quite minimal power because they are sort of relatively small patient numbers. I’ll show you one this was one where I thought we might see some difference. This was in the group of patients who had a rapid increase in creatinine or decrease in GFR in the year before entry.

So these were people who were declining quite quickly but note here I’m sorry that the writing is quite small for those at the back. We’ve got 76 in this analysis, in the latest analysis there are 98 only who actually had that category so 98 out of 800 only had this rapid deterioration. So it’s a small number. You can see that there is a tantalising hint towards revascularisation maybe being better for them but you can see that because the numbers are small confidence intervals are very, very wide and you know we cannot make any conclusion from this. But again I’ll say a little bit more about that later.

Then I’ve presented this data in the States quite a bit over the last 12 moths and we often get the question well that’s great, you take all comers into ASTRAL

but what about those patients who truly have the severest anatomical disease? Well we did have 163 patients who had bilateral RAS greater than 70% or a 70% or more lesion in a solitary kidney with an occlusion on the other side, so a big number of patients and even within them we find by repeated measures analysis this is creatinine there is no difference between the two treatment arms

and when we look at mortality, again there’s a slight hint towards improvement with revascularisation for mortality but again sadly 160 patients only, confidence intervals are too wide to actually prove the point.

We saw the meta-analysis from Liertower and when we go on to consider the complications of ASTRAL there were in fact 7% immediate complications and some of these were fairly serious. We had no immediate mortalities but you can look at 1% getting perforation or dissection of the renal artery, only one patient had a thrombosis as a result of the procedure. Embolisation, stent embolisation.

These are all serious things, access vessel damage so 7% keeping the patients in hospital and then post op complications more than 24 hours, up to 20% of patients had such things as a groin haemorrhage or haematoma or a temporary deterioration in renal function associated with contrast., so a contrast induced AKI. So these procedures are not without their problems and that’s why we’ve got to get the balance right. I think my view is there is no doubt that there is a small subgroup who do benefit but it’s then finding that group as Professor Wilcox has alluded to.

So wrapping up now what are the implications from ATSRAL, from this big study? What can we really conclusively say and what else remains to be said?
Well the first message I would say is that revascularisation is unlikely to improve hypertension or renal function in the vast majority of our patients with atherosclerotic renovascular disease. Those 30.000 patients that are undergoing these treatments per year in the US, the vast majority of those will not improve and I think professor Wilcox’s first slide was showing that maybe only a quarter at best of patients will show some improvement.
In those patients renal artery stenosis is often not pathophysiolgically significant due to many phenomena including the fact that the renal parenchyma is already irretrievably injured from years of hypertension and ischemia and atherosclerotic injury. The second point I would make is and you know I was applying this in my clinics there would seem little point in screening asymptomatic patients. So if you’ve got a patient who just comes to your renal clinic with CKD not necessarily declining just stable CKD and hypertension and you happen to notice they’ve got vascular -- and maybe one kidney a bit smaller than the other at ultrasound is there any point finding renal artery stenosis? Why not just put them on the medication, statins ACE and maybe even anti-platelet therapy? Perhaps pursuing investigation if there is a big dip in renal function with ACE inhibitors.

That would be a non contentious view I would say but I’m emphasising in the asymptomatic patient because there are some groups where we haven’t got the evidence. Then I would point this out to you as well. Notice that in ASTRAL the renovascular patients had a mortality of 8% per year. Now I did this study with Rob Foley who’s based at the USRDS in Minneapolis looking at the Medicare population and we had 1.2 million people in this, 1.2 million who did not have renovascular disease these are all aged over 65 years in Medicare and there were 7000 renovascular patients and back in 2001 the renovascular patients had an annual mortality rate of 16%, whereas those just the elderly Americans without renovascular disease had a mortality of 6.4%.

So what we’re saying is that now 10 years later in ASTRAL the mortality rate of the renovascular patients with high comorbidity is pretty similar to the non-renovascular patients of maybe just under a decade ago. So maybe our medical therapies are improving now as well. This is one implication from that.

Now just to say that yes patients do improve. There are some people who did show an improvement in renal function. What this slide shows is the individual numbers of patients who got benefits in creatinine look of 30µmol/l or even 70 µmol/l benefit after revascularisation. We see it don’t we, we do see the anecdotal cases and look, we had about 8 patients who had a 70 µmol/l or more improvement but notice that in the main we had a similar number of patients who got the same improvement in the conservative therapy arm as well, we shouldn’t forget that.
So my concessions are some patients do benefit from revascularisation. I have no doubt the trick is identifying these. We’ve talked in the last talk about possible ways of trying to identify these. We’re not there yet. None of the patients entering ASTRAL had any functional tests, none of them. In none of them did we measure the pressure gradient across the stenosis and I don’t think even in CORAL that they are now routinely doing that. ASTRAL cannot answer all the questions. One big trial cannot answer every question. We were not powered and we haven’t looked at the impact of revascularisation on ‘flash’ pulmonary oedema upon congestive cardiac failure. As you saw we’re not powered enough in the very, very most severe anatomical group to say can revascularisation improve mortality? We didn’t power the study like that, I think CORAL is trying to do that because it’s being more selective with the entry criteria into CORAL in terms of anatomical disease. We have got 4 year follow up coming up and four subgroup analyses so we might get more information that slightly changes our messages.

So just to summarise what now would I say are the indications for renal stenting in renovascular disease? Let’s now talk about function. I would say if a patient presented to you today with serious acute renal failure and you spotted that there was a serious renal artery stenosis, I would say have a go there’s nothing to be lost and there’s a very good chance that you would improve that patient’s function, get them off dialysis. Is there any evidence for that? The answer is no, but it’s just from the heart we all now that.
What about the old idiom that we would try to preserve renal mass just because there’s an 80% stenosis there? We haven’t got the evidence that there’s any benefit in doing that. ASTRAL would suggest that. In people who are asymptomatic stable CKD, moderate hypertension we have no evidence that intervening in them does any good for them at all.
What about the people with rapid deterioration in renal function in the year before you see them? The evidence isn’t there yet. We’ve got this tantalising hint that there might be some benefit perhaps CORAL may answer it I’m not sure. Maybe a meta-analysis between ourselves STAR and CORAL might help to answer that and get enough patients in that subgroup.

What about heart failure? Flash pulmonary oedema again this is one of those areas that is contentious. Most people don’t like to sit there and see these patients have a mortality risk from heart failure and they would stent. I think I fit into that category but we’ve never had a trial to look at that.
What about congestive cardiac failure? We’ve not had any trial data I think that could be the next trial. Does stenting for heart failure actually improve mortality? An interesting one. What about those patients who need ACE inhibitors but can’t tolerate them? Again we haven’t got the evidence, maybe from CORAL we’ll find out more.

I think this is my last slide. What about hypertension and survival? I would think for severe hypertension ASTRAL would suggest that there’s no benefit in just intervening for that alone. What about patient with accelerated-phase hypertension? A different issue. I think we haven’t got the evidence no one’s tested that in atrial and what about improving survival in the highest risk patients? The answer is we haven’t got it. No one has powered a study to look at this. Maybe CORAL will, if not a meta-analysis between the 3 in bigger studies might do that for us.

So I think I’m going to thank our sponsors and leave my talk there. Thank you.

Chairman: Thank you very much Doctor Kalra we have time for 3 or 4 questions from the audience with the microphone.

Question: More than 70% of your patients have bilateral renal artery stenosis. In how many of them have you done bilateral revascularisation? The second one, do you think that the serum creatinine is a good determinant for following up the patients even with bilateral renal artery stenosis?

Dr. Kalra: I mean that’s a good question so the first question I think was we had about 30% of patients with bilateral renal artery stenosis how many of those were revascularized when they were randomised to that arm? The answer to that is that overall 16% of the revascularized patients had bilateral procedures, so that means half of them doesn’t it really? So half of them got that ok. The second question again sorry?

Dr. Kalra: Serum creatinine as a marker sorry so I mean the point is here it’s a good question, we’re not dealing with the Goldblatt kidney this is the important thing. With atherosclerotic renal artery stenosis the contralateral kidney is not good. So I think it’s a good a marker as any. As you saw, 70% of the patients had GFRs of less than 50 so the creatinine was already elevated and we’re looking at what we all believed before was a progressive disease. So I think yes it’s probably crude in terms of sensitivity but you’ve got to balance that , you need a simple test in a big trial like this if we’d gone for single kidney GFR isotopically we wouldn’t have been able to recruit a fifth of the patients. So you’ve got to be pragmatic. Maybe there’ll be some refinements in the future when we do the next trial with a titre subgroup.

Question: Thank you for this excellent presentation. I got the message that the patient with renal artery atherosclerosis nothing can be done for them nearly except managing the comorbid disease which cause the atherosclerosis because it looks that either they have aberrant kidney damage from either the comorbid diseases or form the hypertension itself which is one of the comorbid conditions. So I got this message from this thing, so just to correct me for fibromuscular dysplasia is the management completely different?

Dr. Kalra: I mean I would say so, I would say that with FMD with a tight stenosis there is something like a 30% chance of actually curing the hypertension and about 2/3 of patients see an improvement in blood pressure control as well. So you get a great response and great kidney. With atherosclerotic disease the kidneys are damaged, even the contralateral kidney is damaged.

Chairman: Well thank you I have juts one remark, we too often forget that renal artery stenosis which is atherosclerotic is in fact a bipolar disease because the coronary arteries are involved in so many cases that mortality for example they don’t die of a kidney disease, they die of a myocardial infarction. Yes we’re treating the kidney but we have to think about the heart as well. Thank you very much