New Year's Eve Poll Results

Happy New Year's Eve! I found the results of last week's poll to be surprising...well, I guess I shouldn't be too shocked to see that Nephrology Fellows won the "hardest working fellow in show business" award, being that this is a nephrology-related blog and individuals have a tendency to view themselves as being excessively overworked. Rather, I was shocked by the paltry showing of Gastroenterology Fellows, who in my opinion have a rather tough job at the three major hospitals I've done post-graduate training at (Penn, Brigham, & Mass General). It just seems like there is an endless supply of GI bleeders, and it's not uncommon that they'll get called in during the middle of the night to do a procedure. While Nephrology fellows get called in for dialysis, for the most part the labor involved should be relatively predictable (evaluate patient for need for dialysis, consent, place line, begin dialysis) compared to doing emergent procedures. In any case: most fellows work pretty hard, and my guess is that there is probably a greater variability amongst fellows within a given subspecialty than there are comparing different subspecialty groups with one another. Happy 2010 to fellows from ALL walks of life.

New poll question to jump-start 2010.

Complications of EPO Use

Well, I had hoped to be more of an active blogger during my holiday vacation this year, but my internet access has been much spottier than desired way out in the boondocks of Western France (not that I expect any renal fellows stuck in the hospital to be sympathetic to my plight in the slightest...) I'll post when I can:

Here's a list of some complications of erythropoietin use:

1. Increased thrombotic events. As the CHOIR study and others have now convincingly shown, elevated hemoglobin targets in CKD/ESRD patients are associated with an increased risk of cardiovascular events (the CHOIR study in particular demonstrated a hazard ratio of 1.34 in CKD patients randomized to a high Hgb target compared to a lower Hgb target, using a composite endpoint of death, MI, CHF requiring hospitalization, and stroke).

2. Hypertension: Patients may even experience intradialytic hypertension after receiving erythropoietin.

3. AV graft thrombosis: again, this correlates with the first point, that elevated Hgb concentrations can result in increased thrombotic events.

4. Pure red cell aplasia: rarely, patients can develop anti-EPO antibodies that result in pure red cell aplasia; this was predominantly associated with a form of EPO (Eprex) not marketed within the U.S.

5. Seizures: via a mechanism that appears uncertain.

6. Progression of diabetic proliferative retinopathy: because EPO has some angiogenic activity, it has been suggested to cause worsening of diabetic retinopathy, with some clinical studies appearing to confirm this hypothesis.

A second look at dilantin

The first time phenytoin was addressed as a blog on this site was on the one year anniversary of the inauguration of the RFN. I'd just like to bring it up again, because a question pertaining to phenytoin and its effect on serum calcium levels recently showed up on the nephrology boards. The aim of the question was to highlight the association between phenytoin and hypocalcemia. Interestingly, phenytoin and a number of other anti-epileptic drugs have been implicated as causes of hypocalcemia as well as other disorders of bone mineral metabolism (for a brief review, refer to Pack et al Cleveland Clinic Journal of Medicine 2004; 71:S2). There are a number of proposed hypotheses to explain the association. To begin with, phenytoin induces the cytochrome P450 system, leading to the catabolism of 25-dihydroxyvitamin D and reducing bioavailable vitamin D levels. Consequently, this can indirectly reduce calcium absorption from the gut and cause hypocalcemia. Phenytoin itself has been shown to directly impair the absorption of intestinal calcium in rats. In fetal rats, phenytoin can also impair the cellular response to PTH; the impaired responsiveness of bone to PTH could result in hypocalcemia. While these proposed mechanisms have mainly been observed in animal models, one should still be aware that there is a link between phenytoin and bone disease and that hypocalcemia is relatively frequently seen in patients chronically taking phenytoin.

Mount Rushmore of Nephrology Announced

The readers have spoken and it is time to announce the results of the hypothetical "Mount Rushmore of Nephrology":  the four Nephrology Pioneers whose contributions have mattered most, and deserve to have their imagery preserved forever on a big slab of rock.  The four highest vote-getters over the week were:

1.  Willem Kolff (inventor of dialysis)

2.  Belding Scribner (access pioneer, founder of 1st outpatient HD unit)

3.  Homer Smith (pioneer of renal physiology)

4.  Joseph Murray (Nobel-winning surgeon who performed 1st kidney transplant)

A few comments on these choices.  First, the top two leading vote-getters were both dialysis-related--which is indicative of the dominant role that dialysis has played in the evolution of nephrology, a field which was early on founded more on issues such as electrolyte abnormalities and medical diagnosis using the urine sediment rather than renal replacement therapy.  Second, I debated as to whether or not to put Joseph Murray on the list--he is, after all, a surgeon rather than a true nephrologist, and to some degree I was hoping that John Merrill would have been the representative for the field of transplant nephrology instead.  Then again, Homer Smith ("Father of Renal Physiology") also may not be considered by some to be a "true nephrologist" since he was not an M.D.--he was really a basic scientist/physiologist. Who knows, perhaps in another 50 years, the list might look different.

Check out the latest poll question on the right.  For those renal fellows working on Christmas Eve/Christmas Day:  may your pager be silent for as long as possible!  

Angiojet as a Cause of AKI??

The Angiojet Rheolytic Thrombectomy System is a medical device currently used to re-establish blood flow in a variety of settings. Briefly, a catheter is introduced into an area of thrombosis and a saline stream is directed towards the thrombus of interest. These saline jets generate a localized low pressure zone via the Bernoulli principle, leading to breaking up of the thrombus. The saline and clot particles are then sucked back into the catheter, avoiding potential embolic complications.

Angiojet has been primarily used for removing thrombus from arteries--for example, in acute myocardial infarction or acute graft thrombosis, for instance. It's also been useful in restoring flow in some dialysis accesses. More recently, however, efforts to use Angiojet in the treatment of venous clots--such as DVT or pulmonary emboli--have been attempted.

At a recent Renal Grand Rounds, one of the fellows presented an interesting case of a patient who developed acute renal failure shortly after undergoing attempted Angiojet therapy for a pulmonary embolus. In addition to a rising creatinine and oliguria, the patient developed red, heme-positive urine with only a few red blood cells, and labs reflecting intravascular hemolysis. Although it is not yet rigorously defined as a cause of acute renal failure, others have reported an association between Angiojet and AKI. It is possible that the Angiojet procedure is more likely to result in hemolysis if thrombectomy is attempted in a large, open space (e.g., for treatment of DVT/PE in a larger vein), as there have generally not been such complications when Angiojet has been used in restoring flow to arteries where the area of thrombus is relatively localized.

It will be interesting to see if there are further reports of this type of AKI with Angiojet.  I purposely put TWO questions marks in the title since I fully admit that much more work needs to be done in order to actually prove causality.  

Differential Diagnosis of Red Urine

Red urine: it's not always hematuria. The first step in evaluating the patient who complains of grossly red urine is to perform a standard urinalysis, focusing on (a) whether or not the dipstick turns heme-positive, and (b) whether or not there are red blood cells visualized on microscopic examination.

We are all fairly familiar with the basic differential diagnosis for heme-positive urine with RBCs visualized on microsopic exam: the presence of dysmorphic RBCs and casts is consistent with glomerular hematuria (e.g., glomerulonephritis) while non-dysmorphic RBCs warrants a work-up for lower-tract hematuria (e.g., nephrolithiasis, bladder cancer, etc.).

However, a heme-positive urine specimen in which RBCs are NOT seen on microscopic exam suggests the possibility of either myoglobinuria (e.g., rhabdomyolysis) or hemoglobinuria (e.g., hemolytic anemia, paroxysmal nocturnal hemoglobinuria, etc). A simple way to differentiate between these two possibilities is to centrifuge the patient's blood: the serum fraction will be pink in hemoglobinuria, and clear in myoglobinuria. Myoglobin is a 17kD protein which is rapidly filtered and excreted by the kidney; thus, it should generally not be present in high abundance in the serum. In contrast, hemoglobin exists as a tetramer of 69kD which is bound to haptoglobin, thereby restricting its filtration at the glomerulus and causing the serum fraction to be pink-tinged. Other lab values (elevated CK in the tens of thousands in rhabdo; positive Coomb's test and low haptoglobin in hemolytic anemias) can also be instrumental.

The differential for heme-negative red urine includes some more esoteric diagnoses: acute porphyria, ingestion of beets or certain food colorants, drugs (e.g., rifampin, pyridium, doxorubicin).

Cardiovascular Complications of Kidney Disease

Question: Who is the most important medical subspecialist for a patient with CKD to have regular contact with? Is it the nephrologist? A case could be made, relying purely on mortality statistics, that it's actually the cardiologist. CKD/ESRD leads to a wide range of cardiovascular complications. A quick review:

1. Hypertension: it is rare to find the ESRD patient WITHOUT hypertension. Along with a high rate of essential hypertension afflicted the general Western population, ESRD patients are at the additional disadvantage of having to deal with hypervolemia and sodium retention, which can made BP very difficult to control. Common ESRD medications such as EPO is also associated with high blood pressure, as is secondary hyperparathyroidism.

2. CHF: Both systolic and diastolic dysfunction are commonly associated with CKD/ESRD; in particular, there is a very high rate of LVH (both concentric and eccentric LVH). Furthermore, patients with AV fistulas or severe anemia can develop high output CHF.

3. CAD: CAD is exceptionally high in the CKD/ESRD population, though the mechanism of cardiac ischemia is posited by some to be discretely different from the standard "rupture of an atherosclerotic plaque" model which has been studied primarily in non-CKD/ESRD populations. Although serum troponin levels may be slightly elevated in CKD/ESRD patients simply because it is cleared less efficiently, even small elevations of cardiac troponins in these patients are associated with poor outcomes. Many hypotheses have been put forth regarding the increased CV risk associated with dialysis; the currently favored ones include abnormal calcium-phosphate metabolism, chronic inflammatory state, dyslipidemias, the accumulation of hypothetical cardio-toxic uremic toxins, and the high rate of association between CKD and diabetes.

4. Arrhythmias/Sudden Cardiac Death: The dialysis unit is a great place to uncover underlying EP abnormalities: there are non-physiologic fluid shifts, rapid changes in electrolyte concentrations, and the patient is being closely monitored.

5. Pericardial Effusions: Both uremia and dialysis itself can be associated with pericardial effusion. Main intervention: intensify dialysis.

6. Valvular Abnormalities: in particular, valvular calcifications can result from abnormal calcium/phosphate metabolism. These patients tend to do poorly with valvuloplasty procedures.

7. Pulmonary Hypertension: it is important to be aware of pulmonary hypertension in the patient with CKD/ESRD as it may make the patient ineligible for renal transplantation if it is too severe.

Ischemia-Reperfusion Models

The study of AKI/ATN has relied heavily on one particular animal model: the warm ischemia-reflow model (often referred to as "ischemia-reperfusion injury"), in which one of the renal arteries is transiently ligated off for a set period of time while body temperature is maintained, then opened up and allowed to reperfuse the kidney. A recent review in Kidney International by Heyman et al addresses some of the limitations of this model, mostly in terms of differences between the mouse model of ischemia-reperfusion and the typical human AKI/ATN we experience clinically.

Some the important differences between mouse and human AKI: First, while the warm ischemia-reperfusion model tends to initially target the S3 segment of the proximal tubule and typically leads to overt tubular necrosis, in human AKI necrosis is not always present, and when it is tends to be patchy and most commonly affecting the distal nephron (in particular: the medullary thick ascending limb and medullary collecting ducts). Second, in clinical practice it is COLD ischemia which is often the mechanism of injury (e.g., surgical procedures in which the aorta is cross-clamped is often performed in the setting of a lowered core temperature, and donated kidneys are typically stored on ice prior to transplantation), rather than warm ischemia. Overall, the authors conceded that the rodent ischemia-reperfusion model has been invaluable, but caution against using it to explain all aspects of human AKI/ATN.

The notion of whether AKI/ATN occurs primarily in the proximal versus the distal tubule is not merely of academic interest. According to one version of the "distal nephron model", the reduced GFR experienced in response to medullary hypoxia is actually an adaptive response: decreasing the metabolic demands of tubular epithelia would decrease hypoxic injury; in a sense one could think of the nephrons as "hibernating" in a low metabolic state until they sense that the hypoxic insult has been removed and they can resume optimal cellular function. If this is true, it might caution physicians from trying to stimulate GFR in patients with AKI, instead encouraging attempts to limit tubular energy expenditure.

IgA Nephropathy Treatment Poll Results

The results of last week's poll on IgA Nephropathy are in:  90% of individuals correctly identified that the only treatment shown in randomized controlled trials to consistently benefit patients with IgA Nephropathy is ACE-inhibitors and/or ARBs.  The response was pretty divided beyond that, however, with a variable number of individuals (37% in all) opting for a trial of immunosuppression.   In retrospect, I'll admit that the question was not perfectly worded: whether the patient has already been treated with ACE-I/ARB therapy was not really clear.  If the patient is just recently diagnosed with IgA Nephropathy, a case could be made to avoid immunosuppression initially and to see whether or not ACE-I/ARB therapy alone could suppress proteinuria and hypertension to acceptable levels.  According to this review in Brigham & Women's "Nephrology Rounds" from 2007, a trial of steroids or possibly even more aggressive cytotoxic therapy (e.g., cyclophosphamide) could be attempted in an individual with these lab parameters who is not responding to conservative therapy.  A majority of folks said they would use fish oil, a topic which has been covered previously.  

This week's RFN Poll question is a fun one:  let's assume Obama's stimulus package includes in its budget funding for a "Mount Rushmore of Nephrology" to be built.  This gigantic statue will feature the faces of 4 prominent nephrologists etched into the rock face of a majestic mountain, preferably at one of the national parks in the Western United States. Of all the prominent nephrologists, which four would you choose as being the most influential in the field?  Choose from the list of historic figures listed, and feel free to include "write-in votes" under the "comments" section if your favorite nephrologist is not listed.  The list I made is admittedly biased towards physicians who have been practicing recently enough that the subspecialty has actually been in existence (e.g., not including folks like Hippocrates), and conversely I tended not to include the current heavy hitters of Nephrology, reasoning that the jury is still out on these folks.  Realize that this is for fun, and I'm sure I've made several oversights.  

Transplant Glomerulopathy

As acute rejection rates continue to fall, causes of late allograft loss, such as transplant glomerulopathy (TG), become increasingly important. TG should spring to mind when you encounter a renal transplant recipient who develops heavy proteinuria and progressive allograft failure, usually late post-transplant. The clinical presentation overlaps with that of chronic allograft nephropathy, although proteinuria tends to be greater in TG and patients are likely to have a history of donor-specific anti-HLA antibodies. The pathogenesis is believed to relate to the presence of these donor-specific antibodies, which are often often anti-HLA Class II. These may wax and wane in concentration and, as a result, may not be detected on a single assay but repeat testing is usually successful. C4d staining is typically negative.

Various immunosupressive regimens have been tried, but none are known to be effective. Progressive graft failure and return to dialysis is the usual outcome. Finally, the increasing use of protocol biopsy informs us that ultrastructural changes that predate the TG lesion develop within the first months post-transplant and in apparently well-functioning kidneys.

Atypical HUS & Orphan Drugs

Atypical HUS refers to hemolytic uremic syndrome which occurs without E. coli 0157:H7-induced diarrhea as the inciting event. There are many causes of atypical HUS: calcineurin inhibitors, HIV, pneumococcal infection and rare genetic diseases are all on the differential diagnosis. On this latter point (genetic diseases), about 50% of all cases of atypical HUS are caused by mutations in genes which are involved in the complement cascade: specifically, the factor H, factor I, and MCP genes.
Individuals with complement-associated atypical HUS have a relatively high degree of renal problems, and unfortunately there are few therapies other than standard supportive measures which can be used to help treat it. One obvious strategy for treating this disease would be to target the complement cascade. Enter eculizumab (trade name Solaris), a monoclonal antibody against the complement protein C5 (which if you'll recall your complement cascade is at the nexus of both alternative and classical pathways). Already approved for the treatment of other disorders of complement dysregulation (specifically: paroxysmal nocturnal hemoglobinuria), eculizumab was recently granted "orphan status" in an attempt to aid its testing in clinical trials for patients with atypical HUS.

What does "orphan status" mean, exactly? The Orphan Drug Act of 1983 was passed in order to encourage the development of drugs for diseases with a small market, defined as affecting less than 200,000 total individuals in the U.S. The act gives pharmaceutical companies enough incentive to invest in developing drugs for treating rare diseases via granting a prolonged competition-free period and clinical trial tax breaks. There are several nephrology-related diseases which would fall into this category.

Hemoperfusion

The first order of business for today's Renal Fellow Network is to welcome another regular contributor to the mix: nephrology fellow Albert Lam, of Brigham & Women's Hospital. Fresh off the Nephrology boards, Albert will be periodically posting his renal pearls of wisdom for the benefit of renal fellows everywhere.

Today's post is on hemoperfusion. If you've never had experience with hemoperfusion, you're not alone: according to this 2008 Kidney International paper by Tyagi et al, use of hemoperfusion has sharply declined over the past several years, and the indications for its use are infrequent.

What is it, exactly? The hemoperfusion procedure is employed in the treatment of specific intoxications and consists of running a patient's blood through a column containing adsorbent particles to which the toxin binds and are removed from the circulation. These adsorbent particles can either be charcoal (which binds water-soluble drugs) or various polystyrene resins (which bind lipid-soluble molecules). Since dialysis is generally successful at removing most low molecular weight, water-soluble molecules which are not protein-bound, it makes sense that hemoperfusion would be considered only when dialysis is not effective (e.g., high molecular weight, highly-protein bound, large VOD drugs). Some of the drugs traditionally removed via hemoperfusion are theophylline and barbiturates, both of which have fallen out of favor (and thus may account some for the decline in the use of hemoperfusion). Valproic acid compounds can also be effectively removed via hemoperfusion, though dialysis is marginally effective as well. Other factors which have contributed to lower hemoperfusion usage include improvements in dialysis technology (e.g., use of continuous therapies and high-flux membranes) as well as the high expense associated with hemoperfusion cartridges, which cannot be reused. Complications of hemoperfusion include thrombocytopenia, leukopenia, and hypocalcemia.

The differential diagnosis of syndromes involving hypokalemia, metabolic alkalosis, and hypertension

While there are a number of conditions which can cause the combination of hypokalemia and metabolic alkalosis there are a limited number of disease processes which lead to hypokalemia, metabolic alkalosis, AND hypertension. In order to have hypertension in this setting, the pathological disease process must involve increased sodium (and water) reabsorption which leads to volume expansion and elevated blood pressure. The list of these diseases includes:

1. Liddle's syndrome. Autosomal dominant condition caused by a gain-of-function mutation in the epithelial sodium channel (ENaC) which results in increased Na+ reabsorption.

2. Licorice ingestion and the Syndrome of Apparent Mineralocorticoid Excess (SAME). Ingestion of large amounts of licorice (or licorice-containing tobacco or gun) can lead to inhibition of the enzyme 11-beta-hydroxysteroid dehydrogenase, which converts cortisol into the cortisone in aldosterone target tissues (e.g. collecting ducts). Since cortisol has an equal affinity for the mineralocorticoid receptor compared to aldosterone, it would act as the primary mineralocorticoid if it were not converted into the inactive cortisone. The compound in licorice that is responsible for this enzyme inhibitory activity is glycyrrhetinic acid, which also has some mild mineralocorticoid activity. The mechanism is similar in SAME, in which one has mutations in the 11-beta-hydroxysteroid dehydrogenase enzyme that prevent proper conversion of cortisol into cortisone.

3. Renal artery stenosis and renin secreting tumors. Both of these etiologies are the result of elevated production and secretion of renin leading to hyperaldosteronism.

4. Adrenal hyperfunction. This category includes causes of primary hyperaldosteronism, including adrenal adenoma, adrenal hyperplasis, and adrenal carcinoma.

All of these conditions essentially result in or mimic hyperaldosteronism and can be partly differentiated on the basis of the response of the renin-angiotensin-aldosterone system to the disease processes:

Liddle's -- low renin, low aldo
Licorice and SAME -- low renin, low aldo
Renal artery stenosis and renin-secreting tumors -- high renin, high aldo
Adrenal hyperfunction -- low renin, high aldo

Anabolic Steroids as a Cause of Secondary FSGS?

A very interesting article in this month's JASN by Herlitz et al describes a cohort of 10 bodybuilders with chronic kidney disease, making a compelling case that anabolic steroid use is an underrecognized cause of secondary FSGS. The paper is strengthened by the fact that nine of the patients underwent renal biopsy documenting surprisingly aggressive forms of FSGS in many of these patients, including some with the "collapsing" variant of FSGS and others with surprisingly high degrees of interstitial fibrosis and tubular atrophy which are not characteristially seen with secondary FSGS. Perhaps even more convincingly for providing a link between bodybuilding and secondary FSGS, the authors describe a patient whose serum creatinine and proteinuria IMPROVE once the patient's intensive steroid/exercise regimen stops, then gets worse when he returns to this regimen against the advice of his physicians.

There are other health problems already associated with anabolic steroid use--these include gynecomastia, dyslipidemia, testicular atrophy, decreased fertility rates, some forms of hepatotoxicity, neuropsychiatric disorders, and developing a massively enlarged head a la disgraced baseball player Barry Bonds. It seems as if kidney disease can now be added to the list.

As a caveat, however, the authors also point out that there are potential other explanations for how bodybuilding might be linked to FSGS. These individuals were typically on regimens consisting of complicated cocktails of steroids, growth hormone, insulin, protein shakes, diuretics, and other supplements whose content is not carefully regulated. Therefore it remains possible that a substance other than anabolic steroids is the common nephrotoxin amongst these individuals. Furthermore, it's already known that elevated BMI in obese patients can result in secondary FSGS; perhaps their increased lean body mass itself drives the process. High protein diets may expose podocytes to unusually high serum protein levels; could this contribute to toxicity?

One final tidbit I learned from this article: individuals who take creatine supplements (another baseball reference: this is one of the few substances that home run guru Mark McGwire has admitted to taking) may have a falsely-elevated serum creatinine based on the fact that creatine is coverted to creatinine. This limitation can be overcome by measuring a full creatinine clearance in which a 24-hour urine creatinine and simultaneous serum creatinine are used.

DRIVE-ing to the Correct Solution Regarding Anemia Management?

Poll results from last week's question regarding anemia management in the dialysis patient.  Interestingly, there was a fairly strong consensus on this one, with over 75% of individuals electing to give a course of iv iron to the dialysis patient in question, who was anemic despite relatively large EPO doses and a ferritin level of 900.  

The relevant trials to cite here are the DRIVE and DRIVE-II trials, in which anemic dialysis patients with an elevated serum ferritin (between 500-1220 ng/ml) and low transferrin saturation (less than 25%) were randomized to receive either a course of iv iron or not.  After 6 weeks, individuals in the iron group were found to have greater hemoglobin levels than the control group, with no obvious differences in complications between the two groups.  While this result would seemingly indicate that "giving a course of iv iron" would be the correct choice, there are still some reasonable doubters out there.  For instance, this CJASN editorial by Drs. Spiegel and Chertow rationally point out that "the long-term safety of unbridled IV iron administration has never been established."  Although we may find it distasteful as physicians wanted to do something to help our patient correct their anemia, I think a compelling case could also be made for the first answer suggested in the poll--"continue current anemia management."  

Be sure and check out the newest poll question in the right margin!

Low Vitamin D Levels Predict ESRD

Hypovitaminosis D is implicated in a wide variety of disease states, including insulin resistance and diabetes, cancer and cardiovascular disease. A new new study in JASN suggests it may also influence the development of de-novo renal disease.

The investigators integrated data from NHANES and the USRDS database to examine the contribution of low 25 (OH) D levels (less than 15 ng/ml) to the incidence of ESRD over a median follow-up of 9 years. Participants were initially free of kidney disease, with under 0.5% having GFR less than 60ml/min or microalbuminuria at baseline.

Of 13,328 individuals, there were 62 cases of ESRD at follow-up. They observed a 2.6-fold increased risk of ESRD in those with 25 (OH) D levels under 15 ng/ml, after multivariable analysis. Further sub-analysis revealed that almost 60% of the excess risk of ESRD comonly seen in blacks was accounted for by low 25 (OH) D levels in their study.

It is presumed that blacks are predisposed to low vitamin D levels as a consequence of decreased synthesis in the skin due to pigmentation. Being very much at the other end of the skin tone spectrum, I'm not reaching for the multivitamins just yet: I'd like to see some interventional data first.

Nephrotoxic HIV Drugs

I've done a post on renal-relevant HIV drugs previously, but I just heard an interesting case of tenofovir-induced renal failure at our Renal Grand Rounds and thought I'd mention a few interesting tidbits from the presentation.

1. HIV-Associated Nephropathy (a.k.a. "HIVAN"), which refers to the rapidly progressive GN characterized by a collapsing FSGS pathology on biopsy, is becoming less common with improved HAART therapy. Furthermore, HIVAN occurs almost exclusively in individuals with African ancestry, so it can usually be left off the differential diagnosis of non-African heritage patients with HIV and renal failure. Drug toxicities are accounting for a higher percentage of HIV-associated renal problems.

2. The mechanism of tenofovir toxicity is unclear but appears to have the most toxicity at the level of the proximal tubule, as affected individuals will usually exhibit a Fanconi's syndrome (with attendant phosphaturia, glucosuria, aminoaciduria, etc.) in addition to AKI. One thought is that tenofovir causes mitochondrial toxicity within proximal tubule cells specifically because the drug tends to concentrate there as tubular secretion is an important aspect of the drug's metabolism. Thus, patients with any degree of CKD should either avoid the drug altogether or get a dose adjustment; this can often be missed clinically as tenofovir is often given as a "combo pill" with other HAART medications (e.g., Truvada = tenofovir + emtricitabine; Atripla = tenofovir + emtricitabine + efavirenz).

3. The World Health Organization recently announced that tenofovir-based regimens be considered preferred, first-line treatment for HIV in developing countries. Apparently the authors of this report factored into account all the various HAART drug toxicities in their recommendations (the previously-recommended stavudine-based regimens also have significant toxicity in the form of lipodystrophy, lactic acidosis, and peripheral neuropathy), but the idea of using a drug known to cause AKI in countries where serum creatinine monitoring and dialysis is not routinely available could be troubling. This is obviously a complex topic, though, where a thorough risk/benefit analysis will be key.

4. Other drugs commonly used in the HIV population which may cause renal side effects include valacyclovir (associated with TMA), foscarnet (hypocalcemia, AKI), Bactrim (AIN, hyperkalemia), and crixivan (protease inhibitor which can cause nephrolithiasis).

Billy the Kidney Demonstrates Important Principles of Kidney Disease

Last July, for my Dad's 65th Birthday, I managed to somehow convince the Indiana branch of the National Kidney Foundation to loan me their "Billy the Kidney" mascot costume. While it was in my possession, I figured why not get a little creative and use Billy to help illustrate some important manifestations of kidney disease and/or physiology. See if you can guess what each of the following four poses represents (you can scroll down to the bottom for the full answers.)

#1: #2:
#3:
#4:
Answers:
#1 = acute renal failure.
#2 = polycystic kidney disease.
#3 = passing a stone.
#4 = "taking a leak" (in case you can't tell, the vegetable my daughter is handing Billy the Kidney is a leek. Yeah, I know, that's one bad joke.)

Dialysis billing

Why do we write notes in the chart day after day? Why does the chart even exist? In my mind, the main function of the medical chart (and progress notes) is communication: charting allows physicians (and other members of the health care team) to effectively communicate with one another when face-to-face communication is not possible. In recent years, however, it seems as if charting has taken on a different primary function: that of a medicolegal document. Not only does the medical chart serve as the official log for documenting what was and wasn't done correctly (in the event of a lawsuit), but it also serves an important function in billing--in order for a physician to get reimbursed for a particular service, he or she must formally document that service, typically by using key words or phrases, within the written chart. Failure to do so correctly can result not only in a failure to get paid hard-earned money, but can even result in allegations of billing fraud.

Thus, it's important to know the dialysis billing codes. Despite its central importance, the details of billing weren't really emphasized to me during either my residency or my fellowship training. Here are a few of the important billing codes and important billing issues related to nephrologists, focusing primarily on the relevant "CPT" (or "Current Procedural Terminology") codes for dialysis. If there are any seasoned nephrology vets out there, please feel free to make comments or clarifications, as I'm just personally discovering the coding game:

• Inpatient Dialysis Codes. The main issue to be aware of here is that in order to get reimbursed for supervising a dialysis session, the attending physician must be physically present while the patient is undergoing the dialysis procedure. Seeing the patient beforehand and writing out the prescription is not enough--you have to be present while the patient is hooked up to the machine. I have been advised to write something like, "The patient was seen and examined while on dialysis," as well as being careful that your date & time on the note is consistent with the patient's time on dialysis. Obviously, this applies only to attendings and not to fellows. The standard inpatient hemodialysis code is the "90935". If you see a patient in the morning and then their status changes such that they now require dialysis, you can still get credit for the procedure by billing a "90937" (again, as long as the nephrologist is present for part of the dialysis procedure). A separate set of codes ("90945" for single visit and "90947" for multiple visits) is used for other dialysis procedures such as peritoneal dialysis; I believe this code can also be used for coding CVVHD.

• Outpatient Dialysis Codes. A list of the major outpatient dialysis codes can be found here; they are generally submitted monthly: the main ones for adult dialysis are "90960" (for 4 face-to-face visits per month), "90961" (for 2-3 face-to-face visits per month), and "90962" (for a single face-to-face visit per month). Separate codes exist for patients who are dialyzed for only a few days out of the month, or for supervising home hemodialysis ("90966").
  

The process of nephrology billing is sufficiently complex that the Renal Physicians Association offers full-day "Coding and Billing Seminars" around the country. The discussion above does not even go into the intricacies of billing for individual injectable medications or all the important disease-specific codes (e.g., ICD-9 codes for CKD, hypertension, etc) which are routinely encountered in outpatient clinical practice.

Link Between CMV Status & EPO Requirements

Why do some dialysis patients require so much more Epogen than others? Some patients manage to hit their targets with relatively low doses, and may even require "holding" EPO injections to avoid the upper hemoglobin range where excess thromboembolic events occur; other patients struggle to achieve a Hgb greater than 10 despite massive doses of EPO and seemingly adequate iron stores. The variables are manifold, but one common assertion is that individuals with a high degree of chronic inflammation tend to be the ones with the greatest resistance to EPO.

A more novel link is reported in this month's JASN: a manuscript by Betjes et al describes an association between ESRD patients with CMV-positivity and those with EPO resistance. It turns out that individuals infected with CMV have an altered profile of T-cells: they tend to have high percentages of CD4+ T-cells which lack the co-stimulatory molecule CD28, whereas patients who are CMV negative contain very small (less than 5%) numbers of these cells. These CD4+ CD28- cells apparently are very pro-inflammatory, capable of secreting large amounts of IFN-gamma and TNF-alpha, which the authors cite as a plausible mechanism to explain why CMV-positive dialysis patients tended to have higher EPO requirements (12,000 units versus 6,300 units per week) than CMV-negative dialysis patients.

The study is potentially significant in that it implicates a common virus in aspects of the chronic inflammatory state known to be associated with poor outcomes in CKD and ESRD, and even points towards antiviral medications as a potential therapy for preventing some of the cardiovascular complications of ESRD. Of course, the big caveat is that the association does not necessary reflect causality; for instance, it may be possible that "sicker", more chronically-inflamed dialysis patients may simply be more susceptible to acquiring CMV seropositivity.

SGLT Inhibitors

A promising new class of oral hypoglycemic agents for type 2 diabetes, the SGLT inhibitors, are on the horizon. As they work on the kidney, and potentially have renal side effects, it may be helpful to review them.

These agents work by inhibiting the SGLT2 glucose transporter found in the S1 segment of the proximal tubule, responsible for 90% of glucose reabsorption in the kidney. They work by causing dose-dependant renal glycosuria, lowering both plasma glucose levels and insulinemia in animal models. Agents have the suffix -gliflozin, and examples include dapagliflozin and sergliflozin. Here's a recent review.

Interestingly, humans with loss-of-function mutations in SGLT2 live into adulthood, but experience systemic hypotension with markedly elevated renin levels, hypernatremia and metabolic alkalosis. This seems to be driven by renal sodium wasting occurring in association with glycosuria. These phenomena could also possibly occur as side effects of these drugs. Whether this side effect is a blessing (through improved BP and volume control) or a curse (through an adverse interaction with ACE inhibitors) remains to be seen. But with over 20 prototypic agents in the drug development pipeline, we'll definitely be hearing more about them.

Steroids for AIN? Poll Results

Very interesting results from last week's poll, in which the question asked was essentially, "Do you believe that steroids are effective in the treatment of AIN?"  The vote was generally split 50-50; that is, about half of individuals felt that removal of the drug and conservative management was the correct decision and about half of individuals felt that some form of steroids should be used.  

This topic is interesting to me because I have run into several experienced nephrologists, whose opinion I really respect, who are adamant that steroids SHOULD be used in the treatment of AIN, despite the conflicting data.  One of the major difficulties in evaluating this question clinically is the relative lack of randomized controlled trials, as well as the general trend to diagnose AIN based on clinical grounds rather than a confirmatory renal biopsy.  Recently,  a 2008 retrospective analysis in KI by Gonzalez et al demonstrated a correlation between a delay in steroid treatment for AIN and the final serum creatinine, causing the authors to suggest that if steroids are going to be used, they should be used relatively early in the disease course.

Check out the latest RFN Poll of the Week in the right margin!

Opiates in advanced renal failure

I was taught the rule of 6's when it came to using opiates in dialysis patients:

• they're metabolized by hepatic glucuronidase to 6-glucurinides (e.g. morphine 6 glucuronide or M6G)
• these metabolites are renally excreted, increasing the half-life to 6 times that of the parent compound in ESRD
• peak toxicity may not occur for over 6 hours (or at 6 am while you're on call) as M6G occurs in both hydrophilic and lipophilic forms, and it takes time for the latter to cross the blood-brain barrier and acculmulate in the CSF. Opiate toxicity should be your first thought in any hypoventilating dialysis patient.

The worst offenders are codeine and morphine, and to a lesser extent oxycodone and hydromorphone. Not all opiates behave like this though, and the following three are probably safest:

• Fentanyl has no active metabolites and appears to be well tolerated.
• Tramadol is a non-narcotic analgesic with some effects at the mu opioid receptor. It's usual half-life is 5 hours, increasing to ten in ESRD. Max. dose of 50mg bid. Watch out for serotonin syndrome with SSRI's
• Methadone does not appear to accumulate at all in renal failure, as observed in drug addicts with CKD and ESRD, and appears safe to use.

Finally, the same pharmokinetic properties that render these 3 agents "safe" in ESRD also make them un-dialyzable, so they should be up-titrated cautiously.

Restless Legs Syndrome in Dialysis Patients

Restless legs syndrome (RLS) is one of the potential neurologic complications associated with CKD or ESRD, though it can also occur independently of kidney dysfunction, with some cases of familial transmission documented (though no gene has yet to be identified). The 2003 NIH criteria for the diagnosis of RLS are the following:

(1) the urge to move the legs, usually with unpleasant sensations.
(2) the appearance of symptoms during inactivity or rest.
(3) relief of symptoms with movement.
(4) worsening of symptoms in the evening or at night.

The mechanism of RLS is unclear, but many signs point to (a) the dopaminergic system (dopamine agonists, such as ropinirole, have been observed to successfully treat some individuals with RLS), and (b) iron status (iron supplementation also appears to have some benefit in treating RLS in individuals with depleted iron stores--a problem which may be accentuated in dialysis patients). In addition, there is some suggestion that inadequate dialysis may be a cause of RLS; this study by Kim et al showed that pre-dialysis BUN is associated with RLS symptom severity.

Thus the workup for dialysis patients with RLS symptoms should include an assessment of dialysis adequacy, review of anemia/iron labs, review of the medication list for drugs which have been associated with RLS (certain antidepressants and antipsychotics have been implicated), and any evidence of a positive family history of RLS.

Thiazide Diuretics

There is a good review article in this month's NEJM entitled "Use of Diuretics in Patients with Hypertension" by Ernst and Moser which focuses predominantly on the thiazide class of diuretics. The thiazides are considered a FUNCTIONAL rather than a STRUCTURAL class of medications--that is, not all compounds considered thiazides have the same chemical backbone, but rather they are classified as such based on their common ability to inhibit the sodium-chloride electroneutral transporter in the distal convoluted tubule. The NEJM article breaks down thiazide diuretics into "thiazide-type" diuretics (which are structurally similar, containing a benzothiadiazine group--these include HCTZ and chlorothiazide, for instance) and the "thiazide-like" diuretics (which do not have the benzothiadiazine group--these include chlorthalidone, metolazone, and indapamide).

The hypertension lowering properties of thiazides occur via two mechanisms: an early anti-hypertensive effect is observed as a result of volume depletion, while long-term benefits on blood pressure appear to be due to decreasing vascular resistance, via a mechanism which is not entirely clear.

A few other tidbits about specific thiazide diuretics which may be useful clinically: the half-life of HCTZ is only 9-10 hours; thus, if you really want to inhibit Na reabsorption in the distal nephron it is necessary to dose bid (the traditionally anti-hypertensive dosing is typically given just once daily). Chlorthalidone is unique amongst the thiazides in that it has a very long half-life which enables once daily dosing. Most thiazide diuretics tend to markedly decrease in efficacy at low GFR's (e.g., less than 40 ml/min), though the article states that metolazone is more resistant to this effect than other thiazides, explaining its occasional utility in the diuresis of individuals with the cardiorenal syndrome.

Side effects of thiazide diuretics include hypokalemia, hypovolemia (e.g., orthostatic hypotension), gout flares, a small but measurable increased susceptibility to the development of diabetes mellitus, and rarely causing an allergic reaction to individuals with a severe sulfa allergy.

Happy Thanksgiving

One of the things I was definitely grateful for at this year's Thanksgiving table: I'm not on-call and can spend the holiday with my family! Best wishes to those renal fellows working hard on this holiday.

Poll Results from last week: there certainly appears to be some dissatisfaction with the ABIM Boards exam, with nobody rating it as a perfect test and most feeling that there is at least some room for major improvement. I wonder if the results would be different if the poll was carried out later, after scores come back (I have heard that there is a 90% first-time pass rate...)

New poll on acute interstitial nephritis is on the right...one of the real "gray areas" in nephrology these days.

Transplant Nephrectomy

In my experience, it’s unusual to perform a transplant nephrectomy in a patient who has returned to dialysis in the absence of a ‘hard’ indication such as macroscopic hematuria or severe symptoms suggestive of rejection or infarction. There have been several occasions where I’ve considered transplant nephrectomy for failure to thrive or refractory anemia, but persuading a surgeon that this is necessary can be a challenge. If you’ve ever been in a similar situation, this paper may interest you.

The authors examined all transplant recipients who returned to dialysis over a ten-year period, using the US Renal Data System. This amounted to approx 11,000 people, a third of whom underwent allograft nephrectomy. They found allograft nephrectomy to be independently associated with improved survival; specifically, there was a 32% lower relative risk for all-cause death after adjusting for many factors, including socioeconomic status, comorbidity, donor characteristics and identifiable conditions associated with requirement for allograft nephrectomy. The authors also excluded patients whose transplant lasted less than 3 months, as they would been likely to have an absolute indication for nephrectomy.

Being a retrospective, observational study, one needs to take the results with a pinch of salt, given the possibility of residual confounding and treatment selection bias. Also, there were some important differences in the baseline characteristics between the groups, e.g. patients who received nephrectomy were more likely to be younger, have a higher serum creatinine concentration and a higher serum albumin, suggesting they were better nourished.

This topic strikes me as a relatively straightforward subject for a randomized controlled trial. As it stands, these observations would embolden me to push a little harder for a transplant nephrectomy in someone with a ‘soft’ indication, in the knowledge that it does not appear harmful, and may in fact reduce mortality. And in the dialysis population, that’s no mean feat.

A role for polycystins as blood pressure sensors?

One of the hallmarks of autosomal dominant polycystic kidney disease (ADPKD) is hypertension. True, most patients with advanced kidney disease get hypertension anyways--but often the degree of hypertension in ADPKD patients seems to be especially high. Is there something about the function of the polycystin proteins which explains the high blood pressure?

A recent article by Sharif-Naeini et al in this month's issue of Cell claims a pressure-sensing role for the polycystins. ADPKD is caused by mutations in two genes, PKD1 and PKD2, which encode for membrane proteins termed TRPP1 and TRPP2, respectively. It appears these genes function via regulating local calcium fluxes in and out of the cell. Although most efforts to understand TRPP1 and TRPP2 function have focused on renal tubular epithelial cells (since that's where cysts come from), a variety of extra-renal phenotypes are also observed. In this paper, the authors generated mice deficient in TRPP1 or TRPP2 ONLY in smooth muscle cells. Importantly, they found that the balance of TRPP1 and TRPP2 levels was critical for the ability of arterial smooth muscle to maintain appropriate blood pressure. These results imply that function of the polycystins may be different depending on cell type, and suggest a possible mechanism by which ADPKD patients may be especially prone to the development of abnormalities in blood pressure regulation.

Nephrology Oral History Project

Very cool project: The Nephrology Oral History Project. On their homepage you can access interviews with 14 different nephrologists who practiced during the early days of dialysis, as early as the 1950s and 1960s. It's well-organized in that each interview is broken down into labeled snippets, so you can only listen to the topics that interest you the most.

One of the more fascinating topics in these interviews is reading about how much tinkering and troubleshooting was necessary during the early days of dialysis. Access was a major issue; early on, one would cut down to find an artery and a vein in order to perform dialysis; at the end of the dialysis procedure, the vessels were ligated off and therefore those vessels could not be used again. Needless to say, there are only a limited number of arteries available to suppor dialysis, so other techniques were essential. One nephrologist describes making "home-made" catheters simply by cutting the appropriate length of polyethylene tubing and tapering off the end by holding it over a Bunsen burner. Because they couldn't afford real stainless-steel guidewires, they used guitar strings coated in silicone, which were much cheaper.

So: the next time you grab a catheter kit off the shelf and plug in the ultrasound machine, think about all the additional hassle involved in placing a dialysis catheter our nephrology fellow predecessors had to endure!

Kidney Clips

It's the weekend, I need to decompress from the boards. Here's some random kidney-related clips. The first is pretty bizarre.



This next one is pretty basic, but the animation is good.

Analgesic Nephropathy

I saw an interesting case yesterday. A 71 year old male, CKD III with a syndrome suggestive of chronic tubulointerstitial disease (U/A trace proteinuria only, disproportionate anemia, mild cortical thinning on ultrasound, normal-appearing vasculature on ultrasound). He had longstanding, intermittent flank pain which began in his 50's, ascribed to nephrolithiasis, although no stone was ever seen. Anyway, after some probing, he admitted to taking a drug called Vanquish regularly (i.e. daily for over 20 years). I had to look it up, but it turns out to be a compound analgesic (acetaminophen, aspirin and caffeine). I was surprised, as I thought these compounds were off the market, as they are in Europe. So beware, analgesic nephropathy is alive and well in 2009!

Analgesic nephropathy is a slowly progressive form of kidney disease characterized by polyuria and renal colic (due to papillary ischemia and necrosis) in the early stages, progressing to tubulo-interstitial disease, transitional cell carcinoma and ESRD over many years. It became a rare diagnosis following the removal of phenacetin from analgesic mixtures in the 80's, but still accounted for 10% of ESRD cases in Australia in 1990.
Phenacetin, whose main metabolite is acetaminophen, was originally believed to be the cause. But this was challenged by the appearance of cases where phenacetin was not involved. The modern definition of analgesic nephropathy does not mention phenacetin: "a slowly progressive disease resulting from the daily use for many years of mixtures containing at least two anti-pyretic analgesics and usually caffeine or codeine (or both), which may lead to psychological dependence". The diagnosis is made based on pathgnomic appearances on non-contrast CT (reduced renal size, bumpy contours and papillary calcifications) in the appropriate clinical context.

Poll Results & Post-Boards Weariness

First things first:  last week's poll results showed that the majority of individuals were wary of using physician "Pay For Performance" as a strategy towards increasing AV fistula placement rates.  Proponents of this strategy (see editorial by Hakim and Himmelfarb in a 2009 KI article) would argue that linking AV fistula placement to some type of financial reward (or linking lack of AV fistula placement to some type of financial penalty, depending on your viewpoint) is the surest way to increase the AV fistula rate in the overall dialysis population.  Potential arguments against using a physician "Pay For Performance" strategy would include penalizing doctors who take care of patients with a lower success of achieving working AV fistulas and encouraging the placement of unnecessary AV fistulas, to name a few.  

So I took the boards yesterday, and I'm eager to hear what others thought of the exam.  Fair?  Frustrating?  Piece of cake?  My own view:  it's a hard exam, and not only because it requires 8 hours of concentration--like many of these ABIM exams, there are often 2-3 answers which seem like they could be right; the challenge comes in selecting the "Best Answer" according to whomever wrote the question.  Ever wonder how they come up with the questions?  Here is the ABIM's stated policy on how all their exams are developed.  This week's RFN Poll of the Week is Boards related.  

ASTRAL Trial

Another day, another negative study… this is getting painful! The large, randomized ASTRAL trial comparing renal revascularization to medical therapy found no benefit, and substantial increased risk, in the interventional arm.

806 patients with atherosclerotic reno-vascular disease were randomized to undergo revascularization in addition to receiving medical therapy or to receive medical therapy alone. After 34 months of follow-up, there were no significant differences between the two groups in the rates of renal events, major cardiovascular events, or death. There were some marginal positive outcomes in the intervention group, including better renal function (p = 0.06) and statistically significant reduction in blood pressure medications. However, 23 patients in the revascularization arm experienced serious complications, including 2 deaths and 3 amputations.

ASTRAL has been criticized on a few fronts. First, the trial design is based on the “principle of equipoise”. Essentially, the investigators only enrolled patients in whom they were uncertain as to whether they would derive clinical benefit from revascularization. As such, many patients enrolled in the trial had lesions of dubious clinical significance, as more high-risk patients would not have been randomized. As an example, 40% of those enrolled had less than 70% stenosis, and some patients with stenoses as low as 60% were included.

Personally, I feel a bit sorry for the authors here. They set out to answer the specific question of what to do with the incidentally discovered, moderate-grade renal artery lesion in a CKD patient, and are then criticized for designing a trial that excludes patients in whom intervention is very likely to benefit. Critics of the trial claim that many of the lesions chosen for intervention were of unclear clinical significance, but I believe that was exactly the point. It’s not as if the management approach to such lesions has been firmly established. The bottom line with intervention in renal arterial disease is that we are doing a poor job in identifying the right kidney in the right patient, and there is a lot more work to do. ASTRAL is step in the right direction.

kappa versus lambda light chains in paraproteinemias

Antibodies are comprised of the structure noted on the left:  they have two immunoglobulin heavy chains (in blue) and two immunoglobulin light chains (in green), covalently linked with one another via disulfide bonds (in red).  The light chains can be one of two types, either kappa or lambda; each individual B-cell (which synthesize and secrete clonal immunoglobulin) expresses either kappa or lambda, but not both, for the entire duration of its lifetime.  

Light chains play an important role in several nephrologic diseases.  They are small enough to be filtered at the glomerulus, but in normal conditions are reclaimed in the tubule.  When the resorptive capacity of the tubules are overwhelmed with extremely large amounts of light chain--as is the situation in paraproteinemias--then the light chains may appear in the urine, and potentially even cause damage as in the cast nephropathy.  Furthermore, certain light chains may accumulate in the glomerular basement membrane, causing glomerular disease (light chain deposition disease, LCDD), and still other light chains may be the underlying cause of AL amyloidosis.  

These different patterns of paraprotein-mediated disease tend to be mediated preferentially by either kappa or lambda light chains:

In cast nephropathy, lambda light chain is the most common.

In light chain deposition disease (LCDD), kappa light chain is the most common.

In AL amyloidosis, lambda light chain is the most common.

The typical kappa-to-lambda ratio in the normal human is about 65:35 (about 1:9), and this ratio is often altered in the above conditions.  It is especially important to look at the K:L ratio (rather than absolute values of K and L light chains) in patients with altered renal function, as the decreased GFR will directly lead to elevation in both K and L light chains.   

On a related note, Waldenstrom's macroglobulinemia is also a paraproteinemia, but rarely causes cast nephropathy or AL amyloidosis.  Interestingly, renal damage in this condition may well be caused by hyperviscosity syndrome.  

T-MINUS 1 DAYS TO THE NEPHROLOGY BOARDS.  GOOD LUCK TO ALL!