CME Slides Forum - A. Davenport

A joint Congress by ERA-EDTA and ISN

ACHIEVING BLOOD PRESSURE CONTROL AND THE RISK OF INTRADIALYTIC HYPOTENSION

Andrew Davenport, London, UK

Chair: John T. Daugirdas, Burr Ridge, USA

Vivekanand Jha, Chandigarh, India

davenport

Dr Andrew Davenport
University College London
Center for Nephrology
Royal Free Hospital
London, United Kingdom

Thank you very much Mr Chairman. I’d like to thank the organisers for inviting me to come and talk today.

As Bruce Robertson just showed you for the general population there’s a definition from the Joint National Committee in terms of hypertension and as Bruce also told you from that in the UK the Renal Association that’s the UK Renal Society Standards Committee issued some standards in 2002 and these standards were also taken up by KDOQI later on in 2006 with pre-dialysis blood pressures as he has just told you of less than 140/90 mmHg and a post dialysis blood pressure of less that 130/80 mmHg.

We’ve conducted an audit just to see actually how we are performing in terms of these blood pressures because what you may not be aware of is that in the UK there are league tables of all the dialysis centres and the patients can go on to a free website and look and see how your centre compares with somebody else. So nobody obviously wants to be bottom of the league. What we found was that disappointedly only about 36% of patients met these targets predialysis. About 42% post- and only about a quarter of patients achieved both targets. As you can see in terms of the predialysis blood pressures, they ranged from sort of less than 30% with only 1 centre having 50% achieving target. Post-dialysis blood pressures a little bit better with three of the centres around about the 50th percentile.

So one of the questions has to be when we looked at this and said well, ok what factors are involved in whether patients meet targets or not? The first thing was it didn’t matter whether patients were prescribed blood pressure tablets or not in terms of achieving these targets. There was no particular beneficial effect of one blood pressure tablet such as an ACE inhibitor compared to a beta blocker or calcium channel antagonist and achieving the target. Similarly there’s no difference in terms of whether patients were prescribed one tablet, two, 3 or even more, 4 or more and achieving the target. Then when we looked at the differences between the units, we found no differences, racial differences because London is a very multiethnic society. Similarly blood pressure control or at least I’ll phrase that achieving these targets was no different whether the centre was a very large centre or whether it was relatively small.

So the question arises well why is the achievement so low? Well, sometimes clinicians feel that the target’s too difficult to achieve and don’t try or it maybe they don’t think it’s clinically relevant. The other issue of course, is all these blood pressure measurements we make pre- and post- dialysis are they really representative? I’m going to hear more of that in the next talk.

So one of the key issues with any target is, is it helpful? So here’s this study from the dialysis clinic in the States, there is a large chain of dialysis centres and what happens here is that if the standard, this is looking at KDOQI standards, if a patient achieves the target on this side of the line, then the patient is more likely to survive and therefore the target is of good clinical relevance. As you can see Kt/v, hematocrit, albumin, if you achieve those targets, the patients do better. Calcium, phosphate and PTH not so bad but look you’re right if they achieve the blood pressure target, increased risk of death. So it does really question whether these targets are actually valid.

Now one of the advantages in the National Health Service in the UK is that patients have to wait before they go on the dialysis machine. So we asked Sandy Mitra a couple of years ago to look at blood pressure and he investigated a cohort of 40 patients, he did an average for 40 hour blood pressure measurements and then looked at the blood pressures when they first arrived at the unit, systolic was 158 mmHg, 10 minutes before they dialysed it was down to 150mmHg, then when they started dialysis down to 146mmHg, so you can see almost a kind of white coat type of picture in terms of hypertension. What was interesting was that the blood pressure which probably most represented the 48-hour blood pressure recording was that taken at about 20 minutes after dialysis. Certainly not the predialysis blood pressure or the post-dialysis blood pressure. This has been confirmed by other studies.

This is from the Agarwal group looking at changes in terms of patients measuring their blood pressure with home blood pressure monitoring or ambulatory blood pressure monitoring and showing that in terms of their survival the best prognosis was in terms of a systolic blood pressure 125- 145 mmHg at home, or if they were wearing ambulatory blood pressure monitoring about 115-125 mmHg.

So what I would suggest to you is this, I think blood pressure does matter but it must be measured reliably and then the question then is, ok if blood pressure matters how do we get there? So we have two philosophies, the concept of the Tassin way to dry patients out, remove sodium or to leave patients relatively wet, try and preserve residual renal function and give them blood pressure tablets.

Now Bruce just showed you this slide from the Lancet and what I just want to point out is that if you look at the cardiovascular mortality, we’re down in fact to 5 trials. As you can see, not all the trials showed benefit because obviously this trial doesn’t show benefit, this trial doesn’t show benefit. If we start to look at the trials which do show benefit, there’s the Suzuki trial from the CO-OPERATE data study and I don’t know if any of you know the history behind this trial but you may wish to go and read the correspondence in the American Journal of Nephrology about this trial. One of these other trials 2006 is still in abstract form, not yet been published. The other thing I should say is let’s look at this other trial I’ve put in circled which is an Italian study which again shows a major positive effect of giving blood pressure tablets.

This was giving a beta blocker carvedilol to an ACE inhibitor or ARB in hemodialysis patients with a dilated cardiomiopathy who were in New York classifications of 3 or 4 etc. What you can see here is an improved survival with carvedilol compared to placebo. But these patients were dialysed 4 times a week and they were dialysing with dialysate sodium 144 -150 mmol/l, so slightly unusual.

So let’s go back and say well, what about the Tassin approach? This is also being done by many workers particularly in Turkey and this was recently a study published by Agarwal and his group which basically what he did was he took down patients weight, so this is weight reduction down here and as you can see, as the weight falls the blood pressure comes down. Over a 4 week period they slowly reduced weight by an average of 0.4 kg with a drop in systolic blood pressure.

So if we balance those two up, the Lancet metanalysis over some trials, then you can show by giving blood pressure tablets they reduce the systolic pressure by about 4.5 mmHg. If you take Argaval’s approach to slowly reduce weight, you drop the blood pressure by about 6.9 mmHg. So there are 2 different approaches. If we go back and look at what we did in London, then this is looking at different practices between units.

So units over here on this right hand side of the graph prescribed fewer blood pressure tablets. These units over here prescribed most blood pressure tablets. This is looking at the percentage of patients who achieved the blood pressure targets. Paradoxically you can see here that the unit’s that were best at achieving the targets prescribed fewer antihypertensives. So you can see from this that the paramount view in London is we tend to try and take salt and water away from patients rather necessary than prescribing blood pressure tablets.

So what did we find in terms of determinance of blood pressure? Well, if you go back to these pre- and post-dialysis blood pressures we found there was a relationship between the predialysis systolic pressure, the intradialytic weight gain and the dialysate sodium concentration. Not too surprisingly the higher the dialysate concentration, you had an increased intradialytic weight gain.

So let’s go back and say, ok if we’re trying to control blood pressure and get the blood pressure down, what’s the risk of intradialytic hypotension? So there’s a bind European and North American agreement of a drop in systolic pressure or mean arterial pressure of more than 20 mmHg with clinical symptoms. It’s been reported by various centres as being between 20-30%.

In our own practice we found that hypotension requiring intravenous fluid resuscitation occurred in roughly 1 in 6 treatments. I’m going to show you this one J-shaped curve and here this is diastolic pressures and a high diastolic pressures, this is non-dialysis patients, this is patients looking at incidence of myocardial infarction or stroke showing that at high diastolic pressures there’s an increased risk of both heart disease, myocardial infarction and also stroke. But if you look at the J-shape, although the J goes up, it is for myocardial infarction, it’s not for stroke. So low diastolic pressures because if you think about when does the heart get perfused, it gets perfuse during diastole, so it’s dependent on diastolic perfusion pressures.

So what one can look at is that in the Annals paper patients with coronary artery disease who had a low diastolic pressure is potentially harmful, there’s increased risk of further myocardial infarction. If one looks let’s say at patients with chronic kidney disease, the irbesartan diabetic nephropathy trial, in those patients who are given irbesartan and the diastolic pressure dropped by 10 mmHg or more the relative risk of having a heart attack increased by 61%. So too lower diastolic pressure is not a good idea.

So what happens during dialysis? So here’s a PET scan. So on the left here’s the heart and you can see some colour, some blood supply to the heart and after dialysis the colour seems to have disappeared. So reduction in blood supply to the heart during dialysis. One can look at that in different methods, this is looking at echocardiography, looking at sort of the thickness of the left ventricular wall, predialysis after 4 hours of dialysis you can see large areas where abnormal wall movement and most of this recovers after dialysis but not all.

So one of the questions could be repetitive falls in blood pressure during dialysis to be dropping low diastolic pressures, reduces cardiac perfusion and this maybe the driving force for developing cardiac fibrosis and this concept we see in patients who have been on dialysis for a long time are people with a stiff heart and a low blood pressure. It maybe the consequence of repetitive low blood pressures during dialysis.

Well, first of all let me annul certain myths. We found that prescribing blood pressure tablets did not increase the risk of dropping your blood pressure during dialysis. In addition, we found there was no difference in intradialytic hypotension in those patients prescribed antihypertensives who took them on the day of dialysis compared to those who omitted them prior to dialysis.

Similarly we found no difference in terms of intradialytic hypotension where the patients were prescribed one antihypertensive or 3 or 4 or more, or between classes it didn’t make a difference. But what did make a difference was ultrafiltration and as from Italy has some data that Claudio Ronco very kindly gave me just showing that as the ultrafiltration rate increases then the risk of hypotension during dialysis markedly increases, so most of the problems we’ve got are related to trying to remove fluid in a very short period of time.

Unfortunately, most of us cannot offer prolonged dialysis sessions or over night dialysis or more frequent dialysis to overcome this problem. When we looked at when patients developed hypotension, it’s not too surprising. If you look at the sessions across the week in standard thrice weekly dialysis, more patients become hypotensive on the first dialysis session of the week compared to the last session.

So the take home message from this slide is that if you want an easy life, never walks around the dialysis unit on Mondays or Tuesdays, always go round at the end of the week. Again when we looked at patients who had developed intradialytic hypotension for each of the sessions, their intradialytic weight gain was greatest so these were having to take off more fluid.

Now this is some data from Nathan Levine’s group suggesting that perhaps it’s not just putting on fluid perhaps where it is put. So this is looking at bioimpedence, looking and seeing the amount of the fluid put on around the trunk compared to the leg and it may be that in patients who put on more fluid around the trunk it’s more difficult to take off, more difficult to refill and therefore, perhaps they may be more susceptible to intradialytic hypotension.

The other thing to think about, of course, is vascular tone. Here’s a German study from a few years ago looking at autonomic responses in patients who didn’t have intradialytic hypotension at controls compared to those with intradialytic hypotension showing an impaired autonomic function.

Perhaps in keeping with that we found that if one looked at our diabetic populations compared to our non-diabetics, then for every session of the week the diabetic patient was much more likely to become hypotensive during dialysis than the non-diabetic.

Well ok what can we do about it? Well, what about dialysate sodium profiling? I don’t know if John can see this but we’re quoting one of his studies here Daugirdas et al =7, no benefit. That’s all still true John I hope? Ok. So if one looks at dialysate sodium, there are lots of studies but often the sample size is very small, there’s often no follow up, the models are variable and actually it’s very difficult and you can’t really do metanalysis on these studies. The majority of studies showed a reduction in so-called hypotension and cramps. But in all the studies which then looked at the possibility of sodium accumulation 10 out of the 12 showed that patients became sodium loaded.

When we looked at the dialysate sodiums against intradialytic weight gain, I’m afraid to say as one increases dialysate sodium, the weight gain goes up proportionally.

What else can we do? Well, here’s an Italian study looking at the importance of temperature and cooling patients down. So this was comparing isothermic dialysis, so there’s no change in temperature versus thermoneutral where there’s no change in energy balance. This is looking at a number of hypotensive treatments and you can see here it’s better for patients to be cooled and be isothermic than to have thermoneutral dialysis. These are obviously the changes in systolic pressure and changes in diastolic pressure. So blood pressure was better maintained when patients were cooled with the isothermic method.

As we’re in Italy, here’s a study showing the difference between hemodialysis and hemofiltration, this study is more than 25 years old. What this shows is that it doesn’t matter whether patients are given hemodialysis or hemofiltration, the most important thing is whether they’re cooled. So if you are cooled on hemofiltration or cooled on hemodialysis, your change in blood pressure is much less than if you have warm hemofiltration or warm hemodialysis.

When we looked at our own patient cohorts the practice in London is to cool dialysate. So what we found is that in our patients dialysed against the lower dialysate sodiums there was no increased risk of hypotension because of the cool dialysate. In fact if anything the patients with the highest sodiums because often they weren’t as cooled were more ate risk of developing intradialytic hypotension.

The final point I want to come on to is this. What about using the dialysis machine? Can we use the dialysis machine technology to prevent intradialytic hypotension?

So most machines these days measure a pseudo blood volume by looking at change in hematocrit. The idea being that water is removed from patients therefore, the hematocrit goes up and for the ease of the graph although the hematocrit goes up, the graph goes down with the idea being it’s a fall in volume, the graph goes down. As patients approach their target weight or so-called dry weight, then the rate at which these curves tends to fall gets steeper and when one looks at the recovery when the ultrafiltration is stopped, the recovery is delayed and takes longer.

So the question is whether we can use these sort of systems to try and prevent intradialytic hypotension, a so-called biofeedback controlled hemodialysis. So the idea here is that the patients given a predicted line that’s this sort of blue line here of fall during dialysis.

There is some very nice work from Amsterdam, from Can Ince’s group looking at how blood vessels close down in terms of response to hypovolemia particularly during dialysis. So what actually happens during dialysis is these very small blood vessels which first of all can strip down. Therefore, what’s actually coming back into the central vein is a more dilute blood and therefore, the rise in hematocrit that you would normally expect isn’t there. Therefore, the problem with the biofeedback systems is they always catch up, they’re always behind what’s actually happening in the body. Because these small vessels are the ones that close down first.

So if one looks let’s say this is from the North Italian Co-op group that there was no significant change in mild symptoms of intradialytic hypotension. Even with severe hypotension although the stiffness drops you’re still talking of over 8% of patients. So even these fuzzy systems I’m afraid to say will not be helpful.

So let me sum up. I think blood pressure control of hemodialysis patients is very important. But the issues are these: we need to decide when and how blood pressure should be assessed. This can be addressed in the next talk. We then need to think what blood pressure is acceptable.

Finally, we do need to prevent intradialytic hypotension but unfortunately to do that we really do need to go back to very simple basics and address sodium balance and weight gain.

Chairman: Thank you Professor Davenport. We have time for a couple of questions. Yes

Prof. Davenport: I think the issue is this is that our current practice is that throughout the London area we advise patients to try and stick to a sodium intake of around about 100 mmol/day but like all things one can give advice, we have no idea as to how many patients actually adhere to the advice. There are interesting issues in that as we work in a multiethnic society that there are differences in sodium intake between racial groups but what was of interest was that we found no difference in terms of ‘blood pressure control’ between our patients from the African, AFRO-Caribbean area compared to those from South Asia.

Question: -- I would like to ask you a question. I guess one of the things that concerns me is whether we really need to control blood pressure in these patients and the risks of doing it and especially in terms of residual renal function. It’s indisputable that residual renal function helps patient survival and I think there was some data out of Turkey suggesting that in that group that was aggressively controlled and ultra-filtered their residual renal function fell off more quickly. Do you have any data from your Registry that would perhaps illuminate that concern?

Prof. Davenport: Well, we do have some data and let me first say that’s the Tassin group in Houston also produced an abstract and a poster again showing that in their process of blood pressure control the residual renal function was lost very quickly. So the best -- from the UK we have a couple of centres that practice what’s called top up dialysis i.e. that when patients first start dialysis their residual renal function is calculated and the dialysis prescription is amended accordingly. Those centres which practice that have the longest duration of residual renal function comparable to peritoneal dialysis. Whereas those centres that put patients on and say right you must do 4 hours and that’s it and we’ll take your weight down, most of those centres the average time for residual renal function is in the order of around 6 months. So there is a big difference in terms of practice.

Question: So given that observation why would you even bother trying to advocate? I’m seeing this as a supply side which is lowering dietary sodium which makes one issue but then on the demand side by increased ultrafiltration I’m just concerned this may actually be a dangerous proposition despite all these machines, all these things I had this crazy fellow who just wanted to normalise everybody and everybody was just dropping their blood pressure. I was very concerned about this over aggressive approach in terms of perhaps hurting patients.

Prof. Davenport: I absolutely agree with that but I think the key issue is that in the UK our recurrent guidelines have no clinical targets for blood pressure pre- and post- dialysis because we’ve decided this is an erroneous measurement, it’s unreliable and certainly I take your point that if patients are coming up for dialysis with a high blood pressure and the decision is being made to try and lower that by adding in antihypertensives I don’t think that’s a good idea. Therefore, the vogue in the UK is how can the National Health Service afford to do blood pressure monitoring? Because obviously what we’d like to do would be sort of ambulatory blood pressure or patients measuring blood pressure at home. In a state run system can the state afford that? We certainly feel that pre and post blood pressures simply recorded in the dialysis unit are not an accurate reflection of blood pressure and should not be used to determine certainly antihypertensive prescriptions.