ACUTE RENAL FAILURE IN SEPSIS FROM PATHOPHYSIOLOGY TO THERAPY

Claudio Ronco, Vicenza, Italy

Chair: Ravindra L. Mehta, San Diego, USA

Eric Rondeau, Paris, France

Prof C. Ronco Dept of Nephrology, Dialysis and Transplantation S. Bortolo Hospital Vicenza, Italy

---

Slide 1

Good morning, Mr Chairman, Ladies and Gentlemen. Acute renal failure represents a risk for mortality.

Slide 2

There’s no question that in the past we were prone to think that people were dying with acute renal failure and not by acute renal failure but in recent papers it has been clearly demonstrated that acute renal failure represents an independent risk factor.

Slide 3

And if acute renal failure requires renal replacement therapy then the risk of mortality is multiplied as it is shown in this paper from Metnitz and co-workers.

Slide 4

So, it is very hard to say what the mortality is from acute renal failure because indeed, we may have a spectrum of disease which includes the acute kidney injury plus a series of other organs failing leading to a situation in which multiple organ failure and especially sepsis represent a very complicated syndrome in this case mortality can increase to very, very high levels.

Slide 5

This paper from John Kellum showed clearly that depending on your definition of acute renal failure you may have different mortality rates. So, if we accept this idea, the concept that mortality in acute renal failure has now changed over the years maybe challenged by the fact that indeed, mortality has changed over the years in acute renal failure but unfortunately, we are treating more and more of the mostly severe patients.

Slide 6

Because of this debate a few years ago in Vicenza, Italy took place one of the sessions of acute dialysis quality initiative when we started trying to define what acute renal failure is and in fact, based on GFR or based on urine output criteria the RIFLE categorisation of acute renal failure took place helping people now to define the level of severity of acute renal failure.

Slide 7

This has been validated in several studies and for example, Eric Hoste has presented the fact that if you have an overall mortality of 14.7 and no acute renal failure represents 6.0 when you move to R, I and F your mortality rate increases significantly. So, there is a strict correlation of the RIFLE definition with the mortality.

Slide 8

What are the prognostic factors in acute renal failure that may describe also the situation of the syndrome? Age, altered previous health, hospitalisation before ICU admission, delayed occurrence of acute renal failure, oliguria, severity of illness, sepsis and associated renal dysfunction.

Slide 9

There are some additional risk factors especially in ICU patients that are basically described  by a series  of other organ dysfunctions rhabdomyolysis, coagulation abnormalities and especially bacteraemia, endotoxemia and sepsis with septic shock.

Slide 10

In sepsis there are specific risk factors which are additional to the conditions of acute kidney injury and they are decreased effective volemia, loss of autoregulation at the kidney level, hyper-perfusion and hypoxia in the outer medulla, factors increasing oxygen the band in the thick ascending limb of the nephron, alteration on the intrarenal versus total blood flow ratio, obstruction of capillaries, cytokines released, leukocyte activation, activation and imbalance of vasoconstricting agents and vasodilating agents, nitricoxide and endothelin and also imbalance between pro and anti apoptotic factors and oxidant stress.

Slide 11

So, if we consider the baseline condition in which the patient is admitted to the ICU, sepsis and septic shock and finally, multiple organ failure represent one of the most important issues today to evaluate the patient with acute kidney injury.

Slide 12

In the pathophysiology of acute renal failure from sepsis two main mechanisms have to be considered, one is the change in intrarenal hemodynamics which is again mediated by several of these events that are described there. The second is the presence of inflammatory mediators which lead to a condition of altered endothelial permeability and altered cellular function.

Slide 13

So, we have a balance between the ischemic insult with hemodynamic changes and sepsis. So, with various toxins moving pre-renal damage into renal cell injury.

Slide 14

If we want to go further into detail, we see that the ischemic insult has basically to deal with a condition that most of the time sees the perfusion pressure being reduced and therefore, there is an altered condition of vascolarisation of the kidney. Then we have toxins that may aggravate the condition and in sepsis the systemic inflammation with a cascade of several mediators may contribute to a condition of hypoxia, oxidant stress, general toxicity, altered vasodilation, vasoconstriction, so renal cell injury that goes from a sublethal condition into apoptosis and necrosis. Hopefully because this is an acute condition, regeneration may take place.

Slide 15

What are the potential responses of renal tubular cells to renal injury? First of all cellular dysfunction and loss of polarity and altered viability of the cells is one of the things but in some cases this altered loss of polarity and adhesion molecules may exfoliate viable cells into the tubular lumen. Then we have altered gene expression. The differentiation of cells and finally, lethal injuries which are basically necrosis and apoptosis.

Slide 16

How do cells die from necrosis? They have severe depression of cellular ATP stores, reduced activity of membrane transport pumps, cell swelling, intracellular free calcium that activates phospholipase A2 and a series of intracellular signalling that in the cytosolic medium lead to sublethal damage to the membrane and finally to cell injury.

Slide 17

So, the ischemia-reperfusion model sees the phenomenon of loss of polarity and altered migration of adhesion molecules in the cells leading into necrosis with some places of apoptosis, exfoliation of cells and then if possible, we have due to the integrity of the basement membrane dedifferentiation, proliferation and finally, hopefully the recovery of the situation.

Slide 18

The typical example is this when you see the flattened tubular epithelium, cell exfoliation, intracellular debris and of course, immunohistochemistry can be of great help in describing adhesion molecule alteration.

Slide 19

But the point is if this is true for the so-called ischemia reperfusion model, is this true also for sepsis and septic shock very often where often we see a hyperdynamic state?

Slide 20

Giantomasso and others have shown that very often renal blood flow isn’t decreased in sepsis, as a matter of fact sometimes it is increased.

Slide 21

So, is it possible that from the ischemia reperfusion model we have to move to another model which may include apoptosis? How does apoptosis occur in tubular cells but not in the physiological mode but rather in the pathological mode? Well, you may have physiological activators, the cascade of mediators and intracellular signalling, loss of survival factors and cytotoxic agents that at the end lead to premature apoptosis of the cell.

Slide 22

Especially important is the fact that the cell may receive two types of messages and activate two types of receptors, one is the so-called death receptor pathway which is TNF-mediated, the other is the cell injury pathway, both have  the characteristics of activating a series of cascades through caspases which are pro-activators and also the alteration of the permeability, of the mitochondria releasing cytochrome C activating once again the pro-caspase leading to the activation of the effector capsase which are really the killers for the cell.

Slide 23

So, for example, we know that we have an equilibrium in the Bcl-2 protein family in which we have Bcl-2 anti-apoptotic and bax which are pro-apoptotic. An imbalance between these 2 components of the equilibrium may lead to premature and pathological apoptosis in the cells.

Slide 24

And blocking therefore, bax could be one of the possibilities that may lead to prevention of pathological apoptosis in septic acute renal failure. This is something which has been, for example, explored with the use of activated protein C which indeed has a potent anti-apoptotic effect.

Slide 25

So the model not anymore of the ischemia reperfusion but of the hyperdynamic typical situation of sepsis is that apoptosis is the prevalent mechanism. Now, why is that when you do biopsies you see only limited apoptotic bodies? Because indeed the monocytes and macrophages are rapidly cleaning the scene from these pro-apoptotic and apoptotic bodies and therefore, what remains in most of the cases is a slight appearance of apoptosis and a large appearance of necrosis.

Slide 26

Now, does it mean that therefore, it is important to know in the complex multiple organ failure syndrome whether the primary insult goes into organ dysfunction or into the kidney? Not really because as you can see here, there is a viscous circle that is sustained by what we call today organ cross talk.

Slide 27

The organ cross talk is very well demonstrated by this paper by Kelly that shows that if you make a kidney ischemic within a few minutes you get a jump in IL-1, TNF-alpha and other mediators.

Slide 28

But not only that you have jump in cardiac IL-1, in cardiac TNF after renal ischemia as you can see in this slide.

Slide 29

And the effect is a reduction and impaired cardiac function after renal ischemia.

Slide 30

The interesting aspect about this is that this organ damage far from the kidney can be prevented by blocking, for example, TNF. So there is no question that this is a humoral mediated condition of organ cross talk and multiple organ failure.

Slide 31

At this point, if you want to prevent the situation and you consider RIFLE as a continuum of evolution of the syndrome, you might want to start from the risk analysis reducing the level of injury, preventing further insults and if you need a renal replacement therapy, well at that point what you need is to favour conditions that avoid further insults or simply that lead to renal recovery.

Slide 32

You may want to identify the patient at risk, the time of intervention is crucial, you have to protect the patient from hemodynamic insult, toxic insult reducing nephrotoxicity of required substances like antibiotics and then protecting from sepsis and then doing a renal replacement therapy that really protects the kidney.

Slide 33

In this setting where ischemia is probably one of the final common pathways and the medullary or relative ischemia can be also in hyperdynamic states where you have an intrarenal versus total blood flow ratio alteration, the target is maintaining the renal blood flow. How do you do that? Cardiac output, volume, renal perfusion pressure.

Slide 34

Cardiac output can be maintained with several of the drugs we have but also with a counter pulsation device and ventricular assist devices.

Slide 35

But volume is one of the important things and fluid infusion should be titrated initially to try to achieve the maximum possible cardiac output.

Slide 36

Now, the interesting message about this is the following. If you have a cardiac output which is above 2.5 l/min/square meter, there’s no point in giving more fluid, at that point the patients require vasopressors because vasoplasia is the mechanism. Therefore, you have to watch carefully for pseudo-ARDS which is induced by excessive delivery of fluids.

Slide 37

In a condition where the kidney has lost the typical autoregulation within a certain level of perfusion pressures and if you have a decrease of perfusion pressure renal blood flow decreases, you have to maintain perfusion pressure which remember is the difference between arterial pressure and venous pressure and sometimes in abdominal compartment syndromes you might have an increase in abdominal pressure.

Slide 38

So, the goal is to increase perfusion pressure, precise blood pressure is difficult to identify and it depends on previous conditions and the benefits of vasopressors, of course, have to be balanced against potential adverse effects.

Slide 39

But the important thing is that we nephrologists were extremely afraid about vasoconstrictors, in sepsis we should not, in septic shock we should not because norepinephrin indeed may increase blood flow instead of decreasing blood flow.

Slide 40

Nephroprotective drugs, oh they’re all very nice but they don’t work. Unfortunately the timing of intervention is describing a dichotomy between the results obtained in the lab or in the animal and the results obtained in vivo.

Slide 41

So, when prevention doesn’t work, you have to go for renal replacement therapy.

Slide 42

Now, when you go for renal replacement therapy, you have to balance between the risks related to the disease and the syndrome and the risk related to the treatment. So how aggressive should we be? How early should we start? This is something that is still under debate.

Slide 43

Nevertheless, the clinical scenario is completely different from what we are used to seeing, in the renal ward the patient is compromised, is complicated, is unstable, is mechanically ventilated, it has multiple organ syndrome, so we need an adequate therapy to treat this patient.

Slide 44

Our risk in treating the patient is to be inadequate, to give a low dose, to cause heterogenic hypovolemia, further ischemic insult of the kidney, prolong acute renal failure and increase proinflammatory effect and finally, cause bleeding, thrombocytopenia or hypoxemia.

Slide 45

This study from Gettings clearly showed that if you start earlier probably you have a higher chance of survival of your patients although there are other studies that show that preventive extracorporeal therapies have no effect so you have to select the right window of intervention.

Slide 46

What is the target of your intervention? Well, you want to mimic the kidney, so you want to have blood purification, you want to restore homeostasis, you want to avoid complications and support conditions favouring recovery.

Slide 47

Can you do that with peritoneal dialysis? I think it has been clearly shown with a recent randomised controlled trial that PD induces increased mortality for several complications, however especially in children this might be still an option at least to be considered.

Slide 48

Intermittent haemodialysis, hemodiafiltration even with the most sophisticated machines, for example, using biofeedback and physiological parameters to treat the patient intermittently we know that we have some problems related to let’s say vascular instability, hemodynamic response to intermittency of therapy.

Slide 49

Then we have continuous therapies which represent also possible limitations because the patient is confined to bed, you need continuous anticoagulation and you have a long-term exposure to the extracorporeal circuit.

Slide 50

However, there is definitely a clinical tolerance which is excellent, you have optimal fluid and uremic control, excellent homeostatic control, you have a continuous clearance as you have in the kidneys.

Slide 51

So, I have shown you that with continuous therapies you can control volume much better than with intermittent. You have stable hemodynamics, you have a stable control of metabolic waste products. You have a stable control of acid base.

Slide 52

And in some studies, although this is still matter of debate, this is a metanalysis provided by John Kellum, continuous renal replacement therapy starts to show a risk reduction for mortality but you have to be careful when you analyse these studies to compare exactly the same severity in patients.

Slide 53

Nevertheless, the use of CRRT is increasing all over the world and across continents there is an increase in the use of continuous renal replacement therapy.

Slide 54

Which indeed, has shown in a study published last year by the group called Best Kidney that the dependency on dialysis when you leave the ICU is much lower in those patients who have done CRRT compared to those patients who have received intermittent renal replacement therapy.

Slide 55

Do we have the technology to do that? Well, this slide answers your question. Yes we have the technology to do that.

Slide 56

And we know from our own studies that if you increase the dose that you deliver, you may have the chance to improve survival especially when you exceed the 35 ml/kg/hour.

Slide 57

Now, the septic patient seems to behave slightly differently.

Slide 58

It seems to take benefit also from a further increase in the renal replacement therapy dose so that the higher is the dose, the higher seems to be the survival. This has led to the idea that as we had dopamine in the past we might have CRRT at renal dose and we might have CRRT at sepsis dose introducing the concept of high volume hemofiltration.

Slide 59

For example, a typical patient treated with high volume therapies improves significantly the PaO2/FiO2 ratio although the balance is 0.

Slide 60

There is no fluid withdrawal from these patients. There must be a removal of substances which have been described by Jean Marc Cavaillon as the storm of cytokines that are over produced in the body and are in excess of biological fluids moving from their original autocrine and paracrine effect into endocrine effects.

Slide 61

So how do I remove these substances today? Some of us have proposed higher ultrafiltration volume, high volume hemofiltration.

Slide 62

This idea moved from a study in 1992 of Grootendorst that showed that if you provide high volume hemofiltration in septic pigs, you could improve hemodynamics and survival of the animals.

Slide 63

Also this paper from Oudemans-van Straaten  showed that observed mortality can be reduced compared to predicted one when you use high volume in a cohort of patients.

Slide 64

In order to solve the problem of feasibility our group have proposed the pulsi volume regimen in which we do CRRT at 2 L/hour during the day except 6-8 hours in which we provide ultrafiltration and the hemofiltration at 85 ml/kg/hour, so very high volume. This has the advantage that in the day time the monitoring is much easier and nurses like it better.

Slide 65

What are the effects? Not only reduced norepinephrin levels but also we have improved TNF production from the cells, improved phagocytosis.

Slide 66

And reduced apoptosis. So we are entering with these therapies into the cell function which is an important issue.

Slide 67

Permeability, high cut-off membrane.

Slide 68

We know that there are studies providing high cut-off membrane that really leads to survival in animal models of sepsis that led us to go ahead and study more these open membranes and you will hear more probably in the months to come.

Slide 69

Finally adsorption using this in combination with plasma filtration therapies.

Slide 70

This is adsorption placing the absorbent directly in contact with the blood.

Slide 71

With a significant reduction of cytokine removal.

Slide 72

This is a sorbent providing adsorption of endotoxins.

Slide 73

Again there is an interesting survival rate compared to controls in a study from Japan.

Slide 74

And finally, CPFA where you filter plasma you circulate through the sorbent and then you reconstitute blood and hemofilter.

Slide 75

Once again the effect is important at the cellular level, monocytes which were paralysed are producing TNF and do phagocytosis in an improved fashion.

Slide 76

What we found was not a significant change in plasma TNF levels but a significant improvement in the capacity of the cell to produce TNF and this was mainly due to the impairment effect produced by the plasma.

Slide 77

Here you see the plasma before and after the sorbent cartridge.

Slide 78

So this led to the hypothesis called Peak concentration that if you inject endotoxins in an animal, you have peaks of substances, pro-inflammatory but also anti-inflammatory that are subsequently appearing on the scene.

Slide 79

If this is true, you might have a condition of immunoinflammation and immune hyporesponsiveness which is dependent on the concentration of these substances.

Slide 80

You might even have a simultaneous appearance of these two conditions in the body.

Slide 81

Now if you have this situation of inflammation and this situation of decreased immunoresponse, you might probably apply drugs specific to this condition but if the system is deactivated completely and simultaneously, you do not know what to do.

Slide 82

Our hypothesis is that continuous therapies are eliminating the peaks in an aspecific way, so either proinflammatory or anti-inflammatory and this is why you don’t see a significant change in concentration throughout the treatment but you might reset at the lower level the immuno-derangement.

Slide 83

And therefore, that’s why when you try to treat one specific mediator you fail because all the others may affect organ function.

Slide 84

But when you see use an aspecific form of cutting out the peaks of both proinflammatory and anti-inflammatory mediators, the syndrome can probably be ameliorated.

Slide 85

So the concept is probably that CRRT are moving towards a different kind of dimensional activity for the organs leading to what we call multiple organ support therapy especially in sepsis supporting the endothelial function. With this I thank you very much.

Chairman: Thank you Doctor Ronco. We shall have the questions later.