APOL1 risk alleles and renal transplant function: time for Mission APOLLO

APOL1 is the newest addition to the list of CKD risk factors- possibly never before in the history of nephrology has a gene associated with such a high odds ratio with disease - above 7 for hypertensive CKD & >10 for FSGS. A gene that confers heterozygous survival advantage when present as a single variant allele, but two variants lead to high CKD risk - APOL1 on chromosome 22. A similar picture of the survival advantage is seen with sickle cell disease. For more on the fascinating story of APOL1 itself, see this open access review, NephMadness coverage from 2015, and the NephJC coverage of a more recent research.

A recent study has shown increased risk of CKD and subsequent ESKD in a population of black kidney donors with high risk APOL1 genotype. High risk is defined as presence of two variants (G1 and G2 – so any combination G1/G1, G1/G2 or G2/G2 is high risk) and was associated with faster progression of renal dysfunction and lower pre- and post-donation eGFR (pre-eGFR 98 versus 108; p= 0.03 and post-eGFR 58 versus 68; p=0.01) over a median follow up of 12 years. The second part of the study involved comparison between these donors and non donors from the CARDIA cohort (Coronary Artery Risk Development in Young Adults), based on APOL1 genotype status. After median 11 years, there was no difference in eGFR decline between the two groups when segregated based on genotype status, indicating that it was the genotype which influenced the eGFR decline and not the donation. Other salient features in the study were:

• 2 donors in the high risk group developed ESRD at 10 and 18 years of donation (11 %, but there were only 19 in this group).
• 78% of the donors were first degree relatives of the recipients. So the fact that a substantial percentage of donors were at risk of subsequent renal disease in this study does raises some valid concerns. Should all black donors be genotyped before transplantation?

An original article published in the Annals of Surgery recently studied eligible kidney donors (n=3438) from the CARDIA cohort of 1985-86 and deduced some risk scores based on the clinical and genetic profile of this population. They projected the 25 year pre-donation risk of CKD (eGFR <60ml/min) or microalbuminuria or macroalbuminuria in an 18 year old and a 30 year old potential donor. This risk is for people with no clinical risk factors (family history/pre-hypertension/diabetes). The risk increases significantly if any of these risk factors are present. See the table for some examples (EA – European American; AA – African American):

How about the risk to the recipients? Case reports have shown development of post transplant FSGS in siblings and monozygotic twins, both in the donor and in the recipient. Data from other studies have shown increased risk of allograft failure in recipients if the donor has a high risk genotype. All these data clearly point to the risk of CKD and subsequent ESKD in AA individuals (both donors & recipients) following transplantation. This brings us back to the most pertinent question, should people of African ancestry be genotyped for APOL1 prior to transplantation? The present studies are not well equipped to answer this question. Though it may seem that one obviously should screen (‘11% of donors with high risk variants develop ESRD!’ ‘High risk variants increase graft failure in recipients!’), lets pause and consider this. We know that live donor transplantation is the modality that offers the best survival. We also know that minorities, in US as well as in UK and elsewhere, have a lower rate of live donor transplantation (see NephJC coverage of a recent JAMA study and the ATTOM study, as well as the NephMadness coverage from 2017 on disparities in transplantation). So, would genotyping worsen the disparity? Is our fate inextricably written in our genes? Enter APOLLO.

APOLLO, APOL1 Long-Term Kidney Transplantation Outcomes Network, is a prospective study aimed at  genotyping donors and recipients in transplants involving recent African ancestry in United States and monitoring long term follow up. This will shed more light on this question and even might have a decisive influence on the present KDRI (kidney risk donor index), replacing the race column with APOL1 genotype. We await this important piece of work with interest.

Post by Sriram Sriperumbuduri

The Kidney in Diabetes: Not Always Plain Vanilla Diabetic Nephropathy

Diabetic nephropathy (DN) is well recognized by the glomerular basement thickening (GBM), mesangial matrix expansion and formation of nodules in the mesangium - the classic Kimmelsteil Wilson lesions. The very sight of these lesions makes a pathologist confirm the diagnosis of DN. However it`s not always as straightforward as it seems and these typical lesions do not always dominate the microscopic picture.

Crescents always makes us wonder about the presence of underlying immune-mediated disorders. We investigate the patient extensively with antibody profiling, complement work up and other battery of tests. Sometimes a crescent may appear unexpectedly in the biopsy of a diabetic patient. A diabetic patient can have ANCA vasculitis after all. If the clinical picture is suggestive of immune disorder, the work up for these diseases will help in arriving at the diagnosis. Immunofluorescence and electron microscopy are also crucial in this regard.

But what if it is a crescent of non-immune etiology? Also described as superimposed lesions of collapsing glomerulopathy (CG) on DN, it is characterized by proliferation of parietal epithelial cells (PECs). These 'crescents' do not show inflammatory cells and GBM has continuity, without any breaks and there is no fibrin in the Bowman space. CG lesions in DN are associated with an increased rate of progression of disease with earlier onset of ESRD. PEC markers like Claudin-1 show strong positivity in Bowman space in these lesions. In fact the cells in Bowman`s space in these crescents show double positivity with Claudin-1 and Nephrin (a visceral epithelial cell marker) in one study. This indicates de-differentiation of PECs into podocytes, forming cross bridges across the urinary space with the PECs trying to replace the latter as podocytopenia is a prominent feature in DN.

There are no definite treatment recommendations for these lesions in DN. Based on few case reports available, aggressive control of diabetes, hypertension and use of ACEI/ARB`s might help to some extent. But the overall prognosis is grim. This highlights the need for careful assessment of such patients. Diabetic patients may have true crescentic , but not all 'crescents' in diabetes are crescentic glomerulonephritis!

Post by Sriram Sriperumbuduri

Image from http://renalpathologyreview.blogspot.co.uk/2013/04/diabetic-nephropathy-with-crescents.html

Nodular Glomerulosclerosis – beyond diabetic nephropathy

A glomerular nodule, i.e. an acellular hyaline structure, can have varied etiologies. Most commonly we see it in the setting of diabetic nephropathy (DN). In these cases, it posesses all the associated features of DN on light microscopy (LM) with glomerular basement membrane (GBM) thickening, mesangial matrix expansion and arteriolar hyalinosis.  These nodules stain well with PAS & silver stains. Immunofluorescence (IF) shows linear IgG deposits along the GBM & tubular basement membrane (TBM) and occasional IgM & C3 trapped in the sclerotic areas. Electron microscopy (EM) shows similar features.
A differential diagnosis is amyloidosis, associated with enlarged glomeruli but poor staining with PAS and silver stain. The striking feature of this condition is red appearance of nodules on Congo staining with characteristic apple green birefringence under the fluorescent microscopy. In the most common form of amyloidosis, the AL type, IF shows light chain restriction with lambda > kappa predominating. EM has the characteristic amyloid fibrils, 7-12 nm in diameter with indefinite length and random orientation.

Among immune mediated glomerulonephritis (GN), MPGN also presents with nodules on biopsy. LM is highlighted by a proliferative morphology with splitting and duplication of the GBM. Cryoglobulinemic GN is associated with pseudo-thrombi in capillaries, which in fact represent large sub-endothelial deposits. IF shows IgG and C3 deposits in GBM. Cryoglobulinemia is associated with IgM and predominance of kappa deposits (Type 1). EM reveals electron dense deposits in sub endothelial, mesangial and sometimes in sub-epithelial locations.

Monoclonal immunoglobulin deposition disease (MIDD) can shows nodules. They stain with PAS and silver stains (see image) and have refractile, PAS-positive deposits in the TBM too. IF is characterized by linear deposits along GBM & TBM with kappa>lambda deposits in LCDD (light chain variant) and IgG in HCDD (heavy chain). EM shows powdery deposits in inner GBM, outer TBM and in the nodules.

Fibrillary and Immunotactoid GN have diffuse a proliferative GN/MPGN pattern, sometimes with crescents. They stain with PAS & silver as well, a feature they share with DN. IF is positive for polyclonal IgG and C3 in Fibrillary GN. The immunotactoid variant has monoclonal IgG with kappa/lambda chains. EM shows large, randomly arranged fibrils (16-20 nm in diameter) in the former, and parallel arrayed microtubules (20-50 nm diameter) in the latter.

Fibronectin glomerulopathy is characterized by nodules positive with PAS and negative with silver. EM shows sub-endothelial electron dense deposits. Immunohistochemistry staining for fibronectin is diagnostic.

The last differential is Idiopathic Nodular Glomerulosclerosis. This entity resembles DN in all aspects except that patient is non diabetic. It is associated with long standing hypertension and smoking. Smoke contains glycation adducts which form AGEs and through oxidative stress is thought to create pathology similar to diabetes.

Thus, a nodule in the glomerulus has a wide differential with definite need for various stains, IF and EM to establish the final diagnosis. The other key feature here is most of the above mentioned diseases can have a similar clinical presentation in the same age group.

Post by Sriram Sriperumbuduri (Images from Paul Phelan)

LVAD & Kidney Dysfunction: The Chicken & the Egg

This is part 2 of a blog on LVAD for the nephrologist. Part 1 may be found here.

Predicting reversibility of renal dysfunction is a difficult task in the setting of heart disease and has a significant effect on patient prognosis. Patients with low GFR may have their kidney function improved after LVAD implantation due to an increase in kidney perfusion. Or it may continue to be poor due to pre-existing intrinsic kidney disease. 

• Some studies show that patients with low GFR prior to LVAD had initial improvement in creatinine but a gradual decline in GFR over the next several months. Despite this, GFR at one year was still more than the pre-LVAD level. Interestingly patients with normal pre-LVAD GFR had a small but steady decline in kidney function after the implantation. Probable reasons for the late decline in kidney function could due to a measurement bias (less muscle mass at the time of LVAD surgery, so low creatinine levels post LVAD, muscle mass improves with rise in creatinine), hemolysis or RV dysfunction (seen in up to 10 % patients post–LVAD). 
• Another unique possibility is an effect of continuous flow physiology on renal vessel walls (animal studies show arterial smooth muscle hyperplasia, periarteritis and interstitial inflammation & fibrosis; no human data). Bresco et al, demonstrated an unexpected survival analysis based on GFR change in first month of LVAD implantation. They showed that patients who had greater than 88 % rise in GFR within one month had a poor survival rate, followed by patients with any degree of worsening of kidney function after LVAD implantation. The best survival was in patients with modest GFR improvement (22-47%). One possible explanation for this could be that patients who are severely ill prior to LVAD are likely to be those with worse kidney function. Could they then have a larger improvement in kidney perfusion and therefore kidney function, in the early post LVAD period?

Dialysis Post LVAD:

The incidence of AKI varies between studies, ranging from 7 to 56%, possibly due to different definitions of AKI, severity of underlying of cardiac & kidney disease, and has an adverse impact on patient survival post-LVAD. Patient with AKI also have poor bridge to transplantation rates (52.4 %vs 83.5 %).

LVAD implantation in the ESRD population has an even worse prognosis, with one recent study describing a mortality rate of 81.9 % compared to 36.4 % in a non-ESRD group after a median follow up of over 2 years. Does this preclude patients with advanced renal failure from LVAD implantation? Combined heart-kidney transplantation may be a viable option for some, and for those who do not have recovery of kidney function after LVAD-associated AKI.

Dialysis options in patients with LVAD include HD & PD. There are no head to head trials comparing the two. PD appears more attractive in my view as it offers more gentle ultrafiltration, less risk of systemic infections and it`s a home modality which keeps the infection prone LVAD patients away from hospital. Very few case reports of PD in LVAD have been published which described successful dialysis in these patients. A couple of practical aspects about PD in this setting:

• imaging of the abdomen should be done prior to PD catheter insertion to confirm that no element of LVAD is intra peritoneal
• PD catheter exit site should be far away from the drive line exit site.

HD is the most frequently used dialysis modality in this population.  There are case reports of all forms of HD done in these patients-regular HD, CRRT and even home HD. Access is an important issue to be discussed. 

• For catheter insertion, we need to remember not to reverse the anticoagulation in these patients. The person doing the cannulation should be trained to do the intervention in a fully anticoagulated individual. It's always prudent to use fluoroscopy for these procedures as it reduces the risk of complications, like a guide wire inadvertently damaging the VAD pump. It's important to perform the procedure under aseptic conditions and avoid bedside procedures. Tunneled catheters are preferred in this regard as they are associated with less risk of infections. If we are forced to do the cannulation without fluoroscopy guidance, it`s preferable to use a short guide wire (10-12 cm).
• Long term access should be a fistula or graft if possible. There are no direct studies comparing these 2 accesses. While the initial preference was for graft due to concerns of poor fistula maturation with the continuous flow, there are published cases of patients who had successful fistula creation while on LVAD. Of the 6 cases described, 2 of them required balloon assisted maturation and the other 4 had unassisted maturation which were successfully cannulated. The more difficult aspect of long term access is a thrill or bruit cannot be appreciated in these patients. The only way of assessing patency of the access is with the help of Doppler or direct cannulation.
• Very few case series of patients on HD have been described. One of them included 10 patients who underwent 281 sessions of HD after post-LVAD AKI. 15 of these sessions were interrupted with symptomatic hypotension being the reason in 6 instances (3 catheter related blood stream infections; 3 volume related).
• Dialysis centres accepting patients on LVAD for HD should have their nurses and doctors trained to interpret the basic LVAD parameters and make necessary changes to hemodialysis prescription accordingly. Specifics about BP measurement were described in part 1 of this post.

The increasing number of patients with LVADs definitely poses a challenge to the Nephrologist. Learning about the device parameters and monitoring is fast becoming essential for us to manage these patients. More research is warranted to understand the implications of this extraordinary treatment for renal function and provision of renal replacement therapy.

Post by Sriram Sriperumbuduri

LVAD for the Nephrologist-Part 1

Advanced heart failure is associated with very poor survival rates of about 10 % at one year. Treatment options include cardiac transplantation or cardiac assist devices. A left ventricular assist device (LVAD) is a mechanical circulatory support for patients with advanced heart failure. Previously considered only as a bridge to transplant, these are becoming more common as a ‘destination’, and used in patients with end stage heart disease patients who are not heart transplant candidates. Of interest to us, this poor cardiac function in heart failure is often complicated by renal dysfunction, the so-called ‘cardio-renal syndrome’. This blog focuses on the outcomes of renal dysfunction in patients with LVAD and few aspects of renal replacement therapy (RRT) in them. I would discuss this in 2 parts to facilitate coverage of all aspects.

I will start with few basics of LVAD that helps in understanding the intricacies of monitoring in HD patients. LVAD consists of an inflow cannula that connects to the left ventricular apex and outflow cannula that connects to the aorta, ascending or descending. A pump connects these two cannulas. An external device, a system controller displays all the LVAD parameters and helps in adjusting the device settings. The pump connects to the controller via a drive-line/percutaneous lead that tunnels subcutaneously and exits the abdominal wall. The power supply is through two batteries that are worn by the patient all the time with adequate backups at their disposal, just in case (see image).

The anatomy and physiology of LVAD has changed over decades (see image below), from a pulsatile flow(pf)-LVAD which were big and relied on pneumatic compression system to pull in and out blood through vascular system to continuous flow(Cf)-LVAD which are more compact and use a power operated rotatory element for moving the blood. The complex structure, large size and the decreased durability were mainly responsible for pushing the pf-LVAD from the main stream and making them obsolete. They were frequently placed in the peritoneal cavity and patients were quite troubled from the noise produced by this device.

A Cf-LVAD is much smaller and does not have the complex structure associated with its predecessor. They are more durable and are usually placed in the abdominal wall (e.g. Heartmate 2) or in the pericardial space (e.g. Heartmate 3 or HVAD/Heartware). Two subtypes of Cf-LVAD exist depending on the type of rotatory pump. The axial flow pump rotates like a propeller in a pipe where in blood flows parallel to the pump and the centrifugal flow pump is a spinning disk with blades with outflow of blood tangential to the disk. Heartmate 2 and Heartmate 3/HVAD are examples of axial flow & centrifugal flow devices respectively.

A few points about LVAD parameters at this stage:

1) Pump speed- the only parameter in the LVAD that can be adjusted and directly influences the pump flow. Very high pump speeds may have consequences such a) hemolysis and platelet activation (due to shear stress on cells), b) supravalvular thrombosis (Cf-LVAD increase diastolic blood pressure -> decreases trans-aortic valve pressure gradient -> decreases frequency and duration of AV valve opening -> decreases blood flow through AV valve -> stasis of blood in the supravalvular region -> thrombosis), c) Right ventricular (RV) dysfunction (increased LVAD flow -> more unloading of left heart -> left shift of interventricular septum (IVS) -> increased RV cavity size -> impaired mechanics of RV contraction + pull over septal leaflet of tricuspid valve with regurgitation -> RV failure) and in extreme cases can lead to d) suction event (very high flow rates -> increased unloading of LV -> collapse of LV on the inflow cannula), which can be fatal as it decreases the outflow and causes arrthymia (IVS impinges on the cannula).

2) Pump flow – defines the amount of blood flowing through the LVAD pump in a minute. It can be as high as 10 L/minute. It is directly proportional to the pump speed and inversely to the head pressure (defined as pressure difference between the LV cavity and the aorta). A decrease in preload due to vasodilatation (drugs/sepsis) can increase the flow rate. On the contrary, hypovolemia (e.g. more ultrafiltration in dialysis), RV dysfunction, and tamponade decrease preload and subsequently the flow. Hypertension by increasing the afterload has a similar effect.

3) Pulsatility Index (PI) –a dimensionless variable which reflects the contractility of LV and it varies directly according to underlying LV function with low PI indicating worsening of LV function due to either a decrease in preload (again can happen in dialysis) or progression of underlying heart disease or an effect of negative inotropes on the right heart. During dialysis, it`s safe to maintain the PI.

How to monitor pulse and blood pressure in patients on LVAD? Well it`s tricky, as Cf-LVAD has a continuous flow physiology. So the pulse is not felt in more than half of these patients. Definitely scary to see a living person with no palpable pulse! The presence or absence of pulse depends on underlying LV function. And it`s even more difficult to monitor blood pressure. If patient has a radial pulse, we can measure pressure manually or with an automated machine. If pressures cannot be obtained with this, a Terumo device could be used for recording. This device has 2 cuffs- large & small which are more sensitive for low pulse pressure and hypovolemia. If radial pulse is not felt, then the best way to measure blood pressure is with the help of a Doppler. The cuff is tied over the upper arm and brachial artery is localized with Doppler. The cuff is inflated until the Doppler signal is lost. On slow deflation, the pressure at which the first Doppler signal is heard corresponds to the mean arterial pressure (MAP). So in patients with LVAD, we have essentially only one pressure recording, MAP of `x` mm Hg. It`s always advisable to maintain MAP between 70-80 mmHg. High MAP (especially greater than 90 mmHg) is associated with decrease in blood flow through the pump and thrombosis.

ACE inhibitors or ARBs are the drug of choice for hypertension control in these patients. Vasodilators are frequently used if MAP is very high. Negative inotropes especially non dihydropyridine calcium channel blockers need to be used with caution as they can impair RV function.

LVAD use is ever increasing in cardiology due to the increasing burden of heart failure patients. Its use has changed from being a bridge to transplantation to now being a destination therapy. As a greater number of patients with heart failure have renal dysfunction, LVAD use has implications for the nephrologist as well.

Post by Sriram Sriperumbuduri, Nephrology Fellow Ottawa