Slide 1

Thank you. So, I am very excited to be here. I’ve actually never been to Sweden, so I thank you for inviting me. I’m happy to talk to you about a topic that’s of great interest to me. It’s difficult to cover the pathophysiology of pre-eclampsia in 15 minutes and as I was walking around Stockholm yesterday completely tired and dazed, I realized that I spent exactly one hour of travel time for every minute of presentation today. So that was pointed out as not being particularly green on my part. But nevertheless we’re going to try and cover the pathophysiology of pre-eclampsia.

Slide 2

Before we delve deep into the glomerular physiology I want to kind of bring this disease home to the practicing nephrologist and tell you why every nephrologist should have some notion about pre-eclampsia. So what should fascinate the nephrologist about pre-eclampsia is the fact that by definition the kidney is always involved. Pre-eclampsia is defined as an elevated blood pressure and proteinuria and the other conditions like low grade coagulopathy, superimposed HELLP and superimposed eclampsia tend to be more of the bonus round, but by definition the kidney is always involved and if pre-eclampsia complicates 5-10 % of pregnancies worldwide, it is in fact, the most common kidney disease in the world. Our patient population is at particular risk for the development of pre-eclampsia.

Slide 3

In this large retrospective cohort study that was done with health records from Denver they found that compared to normal pregnancy women with a diagnosis of kidney disease had an increased risk of pre-eclampsia even after they controlled for hypertension, anemia, cardiovascular disease and respiratory disease. In diabetic women the rates of pre-eclampsia increase along with the rates of proteinuria and you can see that when women have over 300 mg of albuminuria, the rates in the literature have been documented to be as high as 64%.
In lupus it’s anywhere from 13-22% and in IgA nephropathy the rates are anywhere from 10-20%. In other glomerular based diseases they are too rare and the studies aren’t there to allow us to risk to prognosticate well.

Slide 4

Another area of interest is the fact that  pre-eclampsia is often the first presentation of unrecognized kidney disease. In this study with Japanese women a post partum kidney biopsy was conducted in 86 of them and 20% had unrecognized glomerular based disease. The vast majority of course were IgA nephropathy and the best predictor for unrecognized kidney disease was the onset before 30 weeks gestation.

Slide 5

I have a large practice in Toronto of kidney disease in pregnancy and I’m amazed how many times young women will go through multiple complicated pregnancies before the realization that they have kidney disease comes to light.
This is a little bit of Canadian data but it also shows that finally the incidence of pre-eclampsia is probably increasing over time. There’s a variety of reasons for this. Number one, advanced maternal age. In Canada, the number of 1st time births to women greater than 30 has tripled over the past two decades. An audience like us can understand that. You go to medical school, you finish your residency, you do post doctoral training and then you have your first child. I got pre-eclampsia and I’m sure many women in the audience are going to follow in my footsteps. So my recommendation is actually to blow off the rest of the conference and get pregnant today. No seriously, the Swedish people, they’re tall, they’re beautiful and you’ll decrease your risk of developing pre-eclampsia. There’s a changing attitude respect to chronic illness, it’s no longer appropriate for nephrologists to tell a woman you should not get pregnant. Now we discuss risks and many women are choosing to go ahead with the pregnancy despite significant risk. Finally, reproductive technologies are coming to the forefront and in my clinic you wouldn’t be surprised to find a 40 year old with IgA nephropathy or type I diabetes who is planning IVF.

Slide 6

Alright so let’s start with the pathophysiology of pre-eclampsia.
Pre-eclampsia is an endothelial cell disease and is the disease of choice for those of us who are tired of hearing about the podocyte. It is characterized by GFR depression and proteinuria and compared to a normal capillary loop you can see the very swollen endothelial cells, the swollen endothelial lining, the subendothelial deposition as well as mesangial cell interposition.

Slide 7

So in this study that was done a number of years ago but is really quite a landmark study they used inulin and pH clearance to measure the GFR in women with pre-eclampsia, as well as in gravid controls and they found quite a marked GFR depression that’s actually not usually appreciated clinically. They did not find a difference in renal plasma flow or a significant difference in oncotic pressure, thus through mathematical modelling they determined that the cause for GFR depression and pre-eclampsia was secondary to a depression of the ultrafiltration coefficient.
For those of you who don’t live day-to-day by reminding yourselves about the formula that defines GFR I put it here for you on the bottom. GFR as you recall, is the product of the ultrafiltration coefficient and the pressure’s within the glomeruli, those that promote filtration and those that oppose it and Kf or the ultrafiltration coefficient is the product of hydraulic permeability which is the resistant to flow through the endothelial glomerular basement and podocyte layer as well as the surface area available for filtration.

Slide 8

They did beautiful electromicrograph digitalized tracing of the endothelial fenestrae. What they found and this is a woman that’s just a healthy woman, she’s an allograft renal donor. This one is mild pre-eclampsia and this one is severe pre-eclampsia and you can see the profound loss of the endothelial fenestrae.

Slide 9

And thus they concluded that the GFR depression in pre-eclampsia was secondary to a combination of decreased hydraulic permeability due to the reduced size and density of the endothelial fenestrae and the sub-endothelial fibrinoid accumulation, as well as loss of available filtration surface area due to mesangial cell interposition.

Slide 10

But what’s not easily answered by our understanding of the GFR depression in pre-eclampsia is why these women get proteinuria. As you saw on the original micrograph, the podocyte are the epithelial cell layer tends to be fairly well preserved and proteinuria in this disease has proven to be somewhat of a mystery. Earlier studies done by Doctor LindenHymer utilizing dextran sieving found non-specific size selectivity issues and concluded that perhaps there was a difficulty with charge selectivity in this disease.

Slide 11

We’ve also heard the description of an FSGS-like lesion in women with pre-eclampsia that correlates with the degree of proteinuria and these lesions have been shown to resolve postpartum. However this is an inconsistent finding, it is was mostly in case reports and series underlying kidney disease was not clearly ruled out and the theoretical basis was unclear.

Slide 12

So why then do women with pre-eclampsia often have nephrotic range proteinuria?
This is an interesting study, it’s fairly recent. It’s claim to fame is that it was probably one of the most criticized studies in the obstetrical literature because they actually biopsied healthy women who were pregnant and biopsied women with pre-eclampsia,. It was a Swedish study but you guys have a good RAB. So this study although there was so much controversy a lot of people didn’t pay attention to the scientific findings and in fact, they did have some interesting scientific findings in this paper. They graded the degree of endotheliosis. 0 was no endotheliosis, 1 was less than 20%, 2 was 20-80% and 3 was greater than 80%.

Slide 13

Although they found some levels of endotheliosis among women who didn’t have proteinuria. The vast majority  of the endotheliosis occurred in women who were diagnosed with pre-eclampsia. The degree of endotheliosis correlated quite nicely with cystatin C, which as you know, is a marker for GFR, but did not correlate with the degree of proteinuria.

Slide 14

So that begs the question why exactly then do women with pre-eclampsia get proteinuria? Well, there have been a lot of developments with respect to the pathophysiology of this disease. Pre-eclampsia has frequently been described as the disease of theories. Everybody agrees on this stage of the disease. There is placental ischemia. Why there’s placental ischemia remains an area for debate but basically women whose placentas don’t implant properly go on to develop oxidative stress and then endothelial dysfunction. For years we talked about the hormonal hypothesis of this disease that’s the balance between NO and endothelin but the results did not explain the entire pathophysiology of the disease. So, more recently, much attention has been turned to the angiogenic factors and some to the RAS. The RAS maybe of interest in linking with the future cardiovascular disease that is also noted in this patient population.

Slide 15

So I in no way am going to take you through this entire picture of the VEGF system in this talk.

Slide 16

I describe my understanding as a work in progress but basically VEGF interacts with two high affinity receptors; the Flt-1 receptor, the KDR receptor and placental GF which is released from the placenta during pregnancy interacts with Flt-1 and they are both important for the maintenance of vascular integrity and are involved in angiogenesis.
Through a very elegant  series of experiments, Doctor Annette Karumanchi using microchip gene array expression profiling of placental tissue noted upregulation of a soluble Flt-1 and this soluble Flt-1 can bind to VEGF and placental GF and thereby prevent them from interacting with their endogenous receptors.
He found the serum levels in pre-eclamptic patients were 5 times higher and of course, the correlated unbound VEGF and placental GF was reduced. At approximately 48 hours post delivery this effect started to disappear and the sFlt-1 could no longer be identified.

Slide 17

He isolated human soluble Flt-1 and using an adenovirus vector injected it into rats and found that they developed hypertension, proteinuria, glomerular endotheliosis and focal foot process effacement which is very similar to the biopsy in the human condition that I showed you today.

His experiment had a number of nice controls. At one point he just used the adenovirus vector and he injected it into pregnant and non-pregnant animals and as you can see, they did not develop disease.  sFlt1 which will bind both VEGF and placental GF resulted in hypertension and proteinuria both in the pregnant and in the non-pregnant condition.
Flk-1 which will bind only VEGF resulted in disease only in the non-pregnant animal due to the unopposed action of PlGF in the pregnant animal.

Slide 18

At the level of the glomerulus, we know that VEGF is produced by the renal podocyte, it travels against filtration because it’s small and because the distance isn’t too far to interact with its receptors on the endothelium.

Slide 19

In this nice work that actually came from Toronto Dr Sue Quaggin created a heterozygous knockout for podocyte specific VEGF and these animals developed endotheliosis, loss of fenestration and eventually loss of the podocyte foot processes. The homozygous condition was not compatible with life and although this would seem like a beautiful model for the study of pre-eclampsia she received criticisms because this was a developmental model and there’s no way to know in a woman who already has a undeveloped glomerulus would she behave in the same fashion.

Slide 20

So Doctor Quaggin took it a step further, created another model that was just recently published in the New England Journal of Medicine. This time she let the rats fully develop. So, apparently you could be a fully developed at 24 weeks, if you’re a rodent. Then she used a Tet-On system and targeted gene knockout with a tetracycline derivative of podocyte specific VEGF, but she found the same pathology. So even when the glomeruli were allowed to fully develop and you knockout podocyte specific VEGF, you get a pre-eclamptic lesion. Unfortunately, though it could not be rescued by VEGF-121 because one of the theories of course is if we understand the pathophysiology of this disease, perhaps we can start to offer some treatments for it as well.

Slide 21

But women are not rodents, men maybe but women are not so what human data do we have? With respect to human data, it’s fairly limited. Right now we have Doctor Foster’s lab that has looked at primary cultured and conditionally immortalized human podocytes and in the presence of VEGF they maintain their integrity and when they expose an auto-antibody to VEGF they result in death.  So they have defined this nice autocrine loop between the podocyte and the endothelial cell.

Slide 22

In this rather morbid study that used renal autopsy tissue they compared 7 women with pre-eclampsia to 2 control women who died a traumatic death and they used immunohistochemical stains, nephrin, synaptopodin and podocin and they found markedly reduced nephrin and synaptopodin in the pre-eclamptic tissue and therefore, concluded that sFlt-1 resulted in dysregulation of podocyte foot process proteins.

Slide 23

The same authors took it a step further and looked at podocytes in urine and used podocin, podocalyxin, nephrin and synaptopodin and they found that in the urine of women with pre-eclampsia you can identify  podocytes and that the degree of podocyturia actually did correlate nicely with the degree of proteinuria. So there you have it, pre-eclampsia is a podocyte specific disease.

Slide 24

In addition to sFlt-1 Doctor Karumanchi recently identified a second factor, soluble Endoglin. Endoglin, we know is highly expressed in endothelial cells and in trophoblastic cells. It’s a co-surface co-receptor in the TGF-Beta system. It is able to regulate vascular tone via interactions with eNOS so it is important again in vascular integrity. He used the same techniques of gene profiling of placental tissue and found up-regulation of a soluble form and soluble Endoglin was highest amongst the women who developed HELLP syndrome.

Slide 25

He used the same model but this time combined soluble Endoglin with sFlt-1 and injected it using the same adenovirus technique into rodent models and he found that the two in combination produced the most severe disease. Here you see that both of them give you the highest map, nephrotic range proteinuria and you start to see the trends of HELLP syndrome with low platelets and an elevated AST.

Slide 26

When they sacrificed these animals, they noted severe glomerular endotheliosis, placental infarction, foetal growth restriction, giant cell inflammation, liver ischemia and necrosis, haemolysis and increased microvascular permeability something similar to what you would see in a woman with severe HELPP syndrome.

Slide 27

These angiogenic markers now serve as the future for a potential diagnostic test for pre-eclampsia. I realize this is a busy slide and it’s hard to see but even in normal pregnancy which is the blue line in both these slides over time these angiogenic factors will increase and that maybe part of termination of the pregnancy. But in women with pre-eclampsia these factors increase sooner and are more marked and in fact, with sFlt these factors are up 5-6 weeks before onset of the clinical syndrome and they’re up sooner with soluble endoglin in the neighbourhood of 2 months. So if these markers begin to increase earlier, they can serve as a potential diagnostic test for women before they develop pre-eclampsia.

Slide 28

In fact, when you use them altogether a high sFlt, a low placental GF and a high soluble Endoglin you have a huge predictive value for the development of pre-eclampsia both in pre-term and in term pregnancies.
Now clearly every nephrologist can see the value of this. If you’re taking care of women with severe kidney disease on the ward who develop proteinuria in pregnancy it’s always a question that comes to us might this be exacerbation of lupus nephritis or might this be a presentation of pre-eclampsia. So far the data with respect to using these angiogenic markers in women with kidney disease is very limited. In fact, it’s limited to a study of approximately 12 patients. It’s also hampered by the fact as all these studies will be is it still our best guess of whether they did or did not have pre-eclampsia and we don’t have pathological confirmation that these women have pre-eclampsia.

Slide 29

Further, I think we have a lot to go in terms of understanding proteinuria in our women with kidney disease. We are here time and time again when you see a woman with nephrotic range proteinuria that this could be hyperfiltration of pregnancy. I have a hard time accepting the notion of hyperfiltration of pregnancy and I think we might be looking at a two-hit sort of phenomenon. In other words if you already have a glomerular based kidney disease and as the angiogenic factors of a pregnancy start to rise towards the end it shouldn’t be that surprising that some of these women start to spill significant amounts of proteinuria.

Slide 30

This, of course, is animal data again but it’s meant to sort of illustrate the point that in these genetically altered mice this is a hypertensive one and this is a renin overexpressing mouse that the inhibitor of VEGF resulted in more profound endothelial pathological disease in these models. So that might be similar to a woman who already has diabetes or who already has IgA nephropathy.

Slide 31

Once the placenta is delivered, the lesion of pre-eclampsia resolves rather quickly. This again we used inulin and pH clearance and did mathematical modelling of Kf and in 56 women with pre-eclampsia we found that  Kf almost normalized by 1 month postpartum. Along with that there was an improvement in MAP and an improvement in proteinuria.

Slide 32

So very, very briefly at the end I just want to touch on the renin-angiotensin system. There is as much research going on in the RAS as there is in the angiogenic system. Some of the key findings in the RAS have included homozygous T235 allele results in abnormal spinal artery modelling. If you expose human umbilical vein endothelial cells to hypoxia, it can induce ACE. We have known for some time that despite the activated RAS, women, pregnant women are insensitive to this. This was shown years ago when we were able to infuse angiotensin into pregnant women. Ang (1-7) is now prominent in the pregnancy literature, it is a vasodilator that can counteract some of the vasoconstrictive effects of the RAS. It’s higher in normal pregnancy relative to pre-eclampsia and aminopeptidase A has been shown to be higher in pre-eclampsia.

Slide 33

Much of the most interesting work however has been around the AT1R. Just briefly what has been recently identified is an IgG autoantibody to the AT1R. The assay used to identify is rat neonatal cardiac myocytes. As you know, when you stimulate the AT1R you get vasoconstriction and this can be blocked by losartan.
When they isolate the autoantibodies with vascular smooth muscle, it proves a source of reactive oxygen species, increased NADPH oxidase and NF-kappa B and thus may play a role in endothelial dysfunction.

Slide 34

The autoantibody has been described in other disease states; humoral rejection and cardiomiopathy. It’s been shown to stimulate release of sFlt from immortalized human trophoblast cells. But it’s still not clear how these two may act in conjunction with each other. It’s been shown to stimulate heterodimerization of the AT1 receptor with the bradykinin receptor and that may play a role in enhanced angiotensin sensitivity. It crosses the placenta but knowledge about it is still limited because not everybody has a lab full of cardiac myocytes.

Slide 35

I think some of the most interesting work comes from Doctor Luft’s lab, he has found the AT1 to be upregulated on the decidual or maternal side of the placenta, which is the maternal side of the placenta.

Slide 36

And if you think about it that may prove a link between the development of pre-eclampsia and vascular disease.

Slide 37

If a woman has a normal RAS it’s a leap right now but you could think that perhaps that might be why that placenta didn’t implant properly and that might be why she’s also at risk later in life for cardiovascular disease. The future risk of cardiovascular disease is now pretty well-established. This was one of the first papers that showed it. It showed an increased risk of death in women who had pre-eclampsia and pre-term delivery which in this case was 37 weeks. When they looked at the causes of death, the vast majority it was due to cardiovascular disease.

Slide 38

We have started a prospective cohort in Canada where we’re following these women out at one year and we’re going to continue to follow them for as long as we can ad you can see that there are differences that emerge at 1 year postpartum. They have a higher blood pressure, higher HOMA index, higher total cholesterol and LDL cholesterol and about 20% of them still have microalbuminuria. We use these values to model CV risk.

Slide 39

We also showed that the risk was higher in women with preeclampsia, women with severe onset pre-eclampsia and pre-term pre-eclampsia.

Slide 40

So the take home points today are that pre-eclampsia is something a nephrologist should know about. It’s common in our population of chronic kidney disease. The incidence may very well be on the rise. Insights into pathophysiology are quite literally exploding and you should watch your Pubmed weekly to see what is new and it’s not a disease limited to pregnancy. We have to be aware that about 20% have unrecognized kidney disease and those women need appropriate follow up and right now we don’t have particular recommendations for CV risk modification but common sense can prevail. Thank you.

Slide 41

Chairman: Thank you for that presentation. Any questions relating to this? One question I would just like to ask is how long after delivery should you wait before undertaking a biopsy to determine whether the patient has an underlying glomerular disease or just pre-eclampsia?

Dr. Hladunewich: My preference is to wait at least 3 months if you can because the healing phase is usually shown to take between 3-6 months. In women that are hypertensive there have been abnormalities even at a year but at least 3 months if you can wait.

Chairman: Thank you.