Poor phosphate control, or is it?

One of the commonest calls that one gets as a renal fellow in the dialysis unit concerns poor flows from catheters. This is often related to catheter-related thrombus formation and there are a number of different strategies employed to treat this complication and to prevent recurrence. In recent years, the use of alteplase to improve blood flows has become commonplace. In some patients who need repeated doses of alteplase to maintain catheter patency, locking the catheters with alteplase after dialysis has been shown to be effective in preventing thrombosis. Recently, however, an interesting side-effect of alteplase use in this setting has been reported.

An 83-year-old woman with ESRD was having occasional trouble with her dialysis catheter and had intermittently required alteplase-locking. She had been on a stable dose of PO4 binders and her PO4 was very well controlled. However, at one of her visits, her PO4 was noted to be 10.1 (previously 3.5-5) with no change in her diet and the following month it was 35.7mg/dl. Her labs were taken in the usual fashion, after discarding this first 5mls from the catheter. It was postulated that the alteplase might be interfering with the assay for PO4.

In order to confirm this, the authors withdrew 4 samples in 2ml increments from a catheter that had been locked with alteplase during the previous dialysis session. The PO4 concentration in the 4 samples was 161.5, 9.56, 6.47 and 5.68 mg/dl respectively. They then reviewed the records of a patient who had intermittent problems with hyperphosphatemia over the previous year. His mean PO4 level on days on which the catheter was not locked with alteplase was 6mg/dl compared to 9mg/dl on days with alteplase. In retrospect, his PO4 binders were increased inappropriately at least once in response to a spurious test.



So why did this happen? It turns out that the makers of alteplase use phosphoric acid for pH adjustment during the manufacturing process. Although the quantity is very small, it is enough to contaminate a draw from the catheter if it is not adequately removed beforehand. Spurious hyperphosphatemia has also been described in samples contaminated with heparin.

Journal Club: Cinacalcet

Elevations in serum calcium, phosphorous and PTH have all been associated with increased morbidity and mortality in observational studies of ESRD patients. A variety of agents are available to control these serum parameters including phosphorous binders, intravenous vitamin D and calcimemetics. None unfortunately, have been shown in randomized prospective fashion to impact mortality. Cinacalcet, a calcimimetic works by increasing the sensitivity of the calcium-sensing receptor to extracellular calcium. When calcium binds these receptors on the parathyroid gland PTH release is inhibited.

Additionally, cinacalcet leads to lower levels of serum calcium and phosphorous as opposed to intravenous vitamin D which tends to raise these values. Given cinacalcet moves all the serum values in what would seem be helpful directions a recent observational paper in KI took a look at its impact on mortality.

To do this Block and colleagues looked at 5976 ESRD hemodialysis patients from DaVita who had received cinacalcet and compared them with 19,174 who did not. Outcomes were then checked in USRDS.

All patients had secondary hyperparathyroidism as defined by the use of IV vitamin D at baseline.

A multivariate adjusted analysis showed a 26% reduction in all cause mortality in patients who took cinacalcet compared with those who did not. Cardiovascular mortality was also reduced significantly but interestingly only accounted for 44% and 46% of deaths in the cinacalcet and non-cinacalcet groups respectively. Patients in the cinacalcet group at baseline tended to be younger, have less diabetes, fewer days of hospitalization and higher BMIs and albumin.

Although the authors attempted to adjust for these and other known confounders this observational study has a number of potential caveats:

• Confounding by indication (we don’t know what clinical factors led to cincalcet presciption)
• Lead time bias (patients in the cinacalcet group may have for some reason been earlier in the disease process at baseline and thus appeared to live longer)

It should also be noted that the study was partially funded by Amgen (the maker of the Sensipar) and that the first author is both a consultant and advisor for them.

Given the above this paper will not change my current practice. Hopefully  these results will be supported by the upcoming multicenter randomized placebo controlled EVOLVE trial which is examining the impact of cinacalcet on a primary composite outcome of all cause mortality or non fatal cardiovascular event.

Mild hypophosphatemia: Does it really cause muscle weakness?

I recently saw a patient with proximal muscle weakness and mild hypophosphatemia, and I wondered whether a serum phosphorous level of 2.2 mg/dL could be the cause. We know that moderate hypophosphatemia is associated with skeletal and smooth muscle dysfunction. Acute hypophosphatemia can lead to rhabdomyolysis, especially in those with chronic phosphate depletion (eg Alcoholics). Severe hypophosphatemia is associated with metabolic encephalopathy and symptoms of irritability, paresthesias, and even coma.

A pubmed search specifically looking for muscle effects of mild hypophosphatemia returned one paper. It examined respiratory muscle function in an inpatient population. The small study evaluated 23 hospitalized patients with serum phosphate levels less than 2.5mg/dl and compared them to a control group of 11 normophosphatemic inpatients. Mean age, length of stay, and number of predisposing factors for hypophosphatemia in each group was not statistically different. Mean initial serum phosphate level for the hypophosphatemic group was 1.9 +/- 0.4mg/dl compared to 3.6 +/- 0.5 mg/dl for the normophosphatemic group, with a correlating statistical difference in the measures of respiratory muscle strength. After about 2 days of repletion, the mean serum phosphate level increased in the hypophosphatemic group to 3.5 +/- 0.8 mg/dl with no difference in the respiratory muscle strength measurements after repletion in the two groups.

• 43 % (3/7) of those with phosphate levels 2.1-2.5mg/dl demonstrated muscle weakness,
• 75% (9/12) of those with phosphate levels 1.5-2.0mg/dl demonstrated muscle weakness, and all (4) of those with phosphate levels less than1.5mg/dl demonstrated muscle weakness.

Based on my brief literature search, it appears that even mild hypophosphatemia is associated with skeletal muscle weakness, and that this weakness is quickly reversible with repletion to normal levels.

Mary Lieu M.D.

I recommend some nuts

I have a fresh case with great learning points that I would like to discuss today. So I will continue with the management of sensitized patients on my next blog.

Case: This is a 70 yo female with ESRD from diabetic nephropathy underwent a deceased donor kidney transplant (tx) about 4 months ago. Her peritransplant period was remarkable for delayed graft function and a bx at one week after tx showed ATN without no rejection. Four weeks after transplantation, her creatinine reached a nadir of 1.16 mg/dL. At that time, she was noted to have hypercalcemia (10.4 mg/dL) and a phosphate level below assay detection. She denied any weakness, confusion, dyspnea or myalgia.

Discussion 1: Hypophosphatemia is a common complication after kidney tx, affecting more than 90% of patients. It usually regresses spontaneously by 1 year after successful renal tx. In the general population, it can occur by one or more of three primary mechanisms: (1) inadequate intestinal phosphate absorption, (2) excessive renal phosphate excretion, or (3) rapid redistribution of phosphate from the extracellular fluid into bone or soft tissue.

Patients after kidney tx exhibit mainly an inappropriately increased urinary excretion of PO4. This is thought to be secondary to a disorder in the regulation of tubular reabsorption of PO4. Possible factors include: (1) increased PTH level and activity, (2) increased levels of FGF-23; and (3) side effects from immunosuppressive drugs (both high dose steroids and tacrolimus downregulate phosphate receptors on the proximal tubule). The mechanism of PTH and FGF-23 phosphaturia is through a decrease in Na/PO4 cotransporter expression on the apical brush border membrane.

Case cont.: Based on extremely low phosphate levels, she was started on K-Phos neutral 500 mg three times a day. On a routine visit two weeks later, she was found to have a creatinine rise to 3.86 mg/dL. Her only complaint at that time was diarrhea. She was admitted for hydration and possible repeat kidney bx. Her additional labs on admission included calcium 6.6 mg/dL, PO4 5.3 mg/dL, PTH 515 pg/mL, 25-vitamin D 32 ng/mL and FK level 10.5 ng/mL The kidney allograft biopsy specimen showed sequelae of tubular injury associated with extensive deposition of calcium phosphate within the distal tubules!!! So acute phosphate nephropathy (APN), like the ones recently reported with bowel preps!

Discussion 2: In the general population, PO4 repletion is usually recommended when phosphate levels are below 2 mg/dL. The presence of symptomatic complications may warrant more aggressive repletion with IV PO4 therapy. However, oral PO4 supplementation has been linked to increase prevalence of nephrocalcinosis in kidney tx recipients, nonetheless APN is an uncommon complication. In our case, we believe there were multiple factors that triggered the CaPO4 deposition, including: the oral PO4 supplementation (~1,500 mg daily), the volume depletion from diarrhea, the prior tubular injury from ischemia, the high FK level (vasoconstriction and PO4 wasting) and the hyperparathyroidism with urinary PO4 wasting. No FGF-23 was measured but I would antecipate a high level as well. In combination, these factors led to a high urinary PO4 load to a single kidney, which precipitated with calcium in the tubules. There was a concomitant decrease in serum calcium and rise in Cr, which has been classically reported in the older literature to be associated with calcium phosphate deposition and renal failure (figure). In the absence of specific guidelines for kidney transplant recipients, we recommend a conservative management of hypophosphatemia, starting with dietary changes and applying close monitoring in patients on oral supplem. Vitamin D deficiency should also be addressed if 25-vitamin D levels below 30. Finally, the use of cinacalcet in kidney tx recipients in a small short-term trial demonstrated to correct urinary phosphate wasting by mainly affecting PTH levels suggesting a possible role of this drug in management.

PS. High sources of PO4 are almonds, nuts, meats and eggs. A 12-oz Coca Cola has about 68 mg of PO4 and Fruit punch 123 mg.

Bone Scan for Diagnosis of Calciphylaxis?

Calciphylaxis (also called calcific uremic arteriolopathy) is a syndrome of vascular calcification, thrombosis, and necrosis, occurring almost exclusively in ESRD patients.  The diagnosis is typically achieved by clinical exam--demonstrating painful, necrotic, and sometimes ulcerating lesions usually on the lower extremities. Because there are other clinical entities which can cause similar lesions (e.g., nephrogenic sclerosing dermopathy, cryoglobulinemia, cholesterol emboli, vasculitis, etc), sometimes we need a "gold standard".  Not uncommonly, a skin biopsy is carried out for the definitive diagnosis of calciphylaxis.  

However, there is some danger in biopsy:  many of these patients have a difficult time with wound healing, and it would be preferable to have a non-invasive manner by which to support the diagnosis of calciphylaxis.  A 2002 KI article by Fine and Zacharias provides evidence that a bone scan is fairly sensitive for this diagnosis and could be used in lieu of a biopsy. Their analysis indicated that 34 out of 36 patients with a diagnosis of calciphylaxis had an abnormal bone scan--most commonly reflected as showing increased uptake in the calves, typically in the areas of pain/ulceration (as shown on the left).  Of note, this particular study does not really address how specific a bone scan is at excluding alternative causes of this type of dermopathy, but could still potentially be useful in the right clinical setting. Unfortunately, calciphylaxis still carries with it a very high mortality rate.

Re-Evaluating the Ca x P Product

Heard an excellent presentation at our Renal Grand Rounds today by Dr. Charles O'Neill of Emory University, regarding the process of pathologic calcification in CKD/ESRD patients.

One of the points I took away from this talk is that we need to re-evaluate the role of the "calcium-phosphate product" in clinical medicine.  Presently, the KDOQI guidelines state that nephrologists should attempt to maintain the Ca x P product below 55 in order to minimize pathologic calcification.  I had always been taught that if the Ca x P product exceeds 55, there is a tendency for calcium phosphate to precipitate and deposit within the walls of blood vessels.  Is this true?

This 20007 KI review ("The Fallacy of the calcium-phosphorus product") makes the case that this view is erroneous.  The majority of medial calcification seen in CKD/ESRD patients is in the form of hydroxyapatite, which is much more complex than calcium phosphate; the chemical structure is Ca10(PO4)6(OH)2.  As the formation of hydroxyapatite involves multiple separate steps, it is very unlikely to occur spontaneously. A more likely scenario is that a local balance between specific calcification inhibitors (e.g., pyrophosphate) and activators (e.g., alkaline phosphatase) maintained at a local level determines whether or not pathologic calcification occurs.  Furthermore, experiments in which exogenous calcium and phosphate were added to samples of human plasma  demonstrated that calcium-phosphate precipitation did not occur until the Ca x P product exceeded over 200--a number which is never achieved in human patients.  

Nonetheless, despite these potential flaws in the physiologic rationale for the Ca x P product, there is abundant epidemiologic evidence showing an association between the Ca x P product and cardiovascular mortality.  Perhaps this simply reflects the observation that serum calcium and serum phosphate levels each independently potentially contribute to pathologic calcification.  In any case, the prevailing current approach is to control serum phosphate levels with binders to within the normal range as much as possible while tolerating some degree of mild hypocalcemia.  

the von Kossa Stain for Acute Phosphate Nephropathy

One of the hot topics in Nephrology over the past few years has been the epidemiologic and histopathologic studies suggesting oral sodium phosphate colonoscopy preparations as a cause of acute phosphate nephropathy. As evidence continued to mount, the FDA in December 2008 forbade the over-the-counter sale of oral sodium phosphate products; however, they are still available by prescription under the names "Visicol" and "Osmoprep." These preps are still favored by many gastroenterologists (and patients) based on the fact that it is much easier to take than the more traditional polyethylene glycol-based colonoscopy prep (e.g., "Go-Lytely").

Biopsies of patients with acute phosphate nephropathy tend to show abundant calcium phosphate crystal deposition, mostly within the distal convoluted tubules and collecting ducts, but sometimes also in the interstitium. A good detection method is the use of the von Kossa stain, which stains certain calcium-containing salts such as calcium phosphate a brownish-blackish color.

Lipid-Lowering Effect of Renagel

This has been known for awhile, but it is worth repeating--sevelamer (Renagel), in addition to its ability to act as a phosphate binder, also results in LDL cholesterol-lowering effects.

Two of the early articles (a 1998 Clinical Nephrology study by Wilkes et al, as well as this 1998 NDT paper by Goldberg et al), first noted this unexpected association.  In the former paper, a group of dialysis patients subjected to sevelamer hydrochloride therapy following a phos-binder washout period was noted to have a 35.9% fall in LDL cholesterol with an HDL cholesterol level that remained relatively stable.  This LDL-lowering effect is also present in the newer formulation of Renagel, sevelamer carbonate.  

It is unclear exactly what this lipid-lowering effect may indicate, especially considering the repeated finding that statins appear to be less efficacious in CKD/ESRD patients than in the general population in terms of lowering cardiovascular mortality.  Still, it is tempting to speculate that cholesterol-lowering should be a good thing for CKD/ESRD patients given their high rate of CVD.  The jury is still out in terms of what is the optimal phosphate binder to use in a given clinical scenario, but the lipid-lowering effect of Renagel does make it somewhat unique from the other phosphate binders available.  

Refeeding Syndrome

The clinical entity of refeeding syndrome was first described in U.S. POWs in Japan who had been starved for many months to years, then developed extreme electrolyte abnormalities upon the reintroduction of food.

The most common electrolyte abnormalities associated with refeeding syndrome are hypophosphatemia, hypomagnesemia, and hypokalemia. The mechanism is as follows: during starvation, the secretion of insulin is decreased in response to a reduced availability of carbohydrates, and in order to survive metabolism is shifted to using fat and protein stores. During this adaptation total-body phosphate, magnesium, and potassium stores are depleted, despite maintaining relatively normal serum concentrations of these electrolytes. When suddenly exposed to ample food (and in particular carbohydrates), there is a shift to a carbohydrate-based metabolism associated with a sudden surge in insulin secretion. As we all know, insulin stimulates a rapid intracellular shift of potassium and phosphate, which can lead to a profound decrease in this electrolytes. In particular hypophosphatemia can lead to catastrophic muscle dysfunction, including respiratory collapse, as phosphate is necessary to maintain ATP stores needed for muscle contraction. Hypokalemia is well-known to result in cardiac arrhyhthmias.

Patients with alcoholism, anorexia nervosa, prolonged hospitalizations, or cancer patients may also be subject to refeeding syndrome. It typically occurs within four days of refeeding. Either hyperglycemia or hypoglycemia may also be present. Careful monitoring of electrolytes in patients prone to refeeding syndrome, as well as searching for nutritional deficiencies which may go along with refeeding syndrome (e.g., thiamine deficiency) are the cornerstones of preventing this potentially dangerous complication.

Phosphate Binders Lower Mortality in ESRD Patients

I heard a talk today by one of the fellows in our program who is also the senior author on this recent JASN article: Phosphorus Binders and Survival on Dialysis.

Interestingly, even though the current KDOQI Guidelines recommend the use of phosphate binders to maintain serum phosphorus levels within the normal range, there has never been a randomized controlled trial demonstrating the superiority of their use compared to placebo. The FDA approval of phosphate binders is based primarily on their ability to lower serum phosphorus, rather than any hard outcomes such as mortality or cardiovascular disease.

Furthermore, the practice of prescribing phosphate binders is so ingrained in nephrologists, and is considered the standard-of-care based on the current practice guidelines, that it would be considered ethically inappropriate to conduct a phosphate binder-versus-placebo prospective study today.

As such, the authors of the above study looked at a cohort of over 10,000 patients within the Fresenius dialysis system and divided them into two groups, which were followed prospectively: those who were prescribed phosphate binders, and those who were not. Obviously, it is difficult to completely eliminate confounders in a study like this, but the authors attempted to do so by performing a technique known as propensity score matching, in which patients with similar baseline characteristics are matched with one another and those at the extremes (e.g., too healthy, too sick) are excluded from the study. I'm no stat-man myself, but a friend of mine who is claims that propensity score matching is becoming the accepted standard for these types of epidemiologic studies. In any case, the authors of this study found a substantially decreased one-year mortality in patients who were treated with phosphate binders compared to those who were not.

Sodium Thiosulfate for Calciphylaxis

Calciphylaxis is a rare but severe dermatologic condition affecting end-stage renal disease individuals for reasons which remain obscure. Histologic biopsy is the gold-standard for making the diagnosis; it typically demonstrates calcification within the media of small and medium-sized arterioles with intimal hyperplasia and fibrosis, leading to its other name, calcific uremic arteriolopathy.

Treatment strategies for calciphylaxis remain a challenge for this condition, and for the most part involves supportive care with antibiotics as necessary, pain control, and judicious management of serum calcium and phosphorus levels. One relatively recent potential therapy is sodium thiosulfate. Although its use is still considered "off-label", there have been some encouraging case reports and also animal studies to support its beneficial effect on pathologic calcification.

For those who wish to try sodium thiosulphate infusions, iv doses generally vary from 5-25 grams after or during dialysis over 30-60 minutes. Although generally well-tolerated, GI effects as well as an anion gap metabolic acidosis (due to the thiosulphate moiety) have been reported as side effects.

Phosphate Transporters in the Proximal Tubule

Phosphorus handling is a hot topic in renal physiology, especially considering the advanced being made in bone metabolism suggesting that vascular calcifications may play some role in the high mortality rate of chronic kidney disease patients. But how is it regulated at a molecular level?

The predominant phosphorus uptake channel in the proximal tubule is Napi channel, a Na-phosphorus cotransporter. Its regulation appears to be largely due to regulated trafficking. At basal conditions, the transporter is expressed on the apical surface of proximal tubule cells. However, when the cell receives signals from either PTH or FGF23 (which are secreted in response to elevated phosphorus levels), it will internalize Napi via clathrin-coated pits and it will ultimately be degraded; ultimately this will lead to decreased renal phosphorus reabsorption and a gradual decrease in phosphorus levels.

Acute Phosphate Nephropathy

It has long been known that acute phosphate toxicity--as might be seen in rhabdomyolysis or tumor lysis syndrome--can lead to acute kidney injury as a result of calcium-phosphate deposition in the renal parenchyma and tubules. Interestingly, this problem appears to be increasing as a cause of AKI given the increased use of phosphate-containing bowel preparations (such as Fleet's enema) for screening colonoscopies.

Recent epidemiologic evidence from Khurana et al (Arch Int Medicine, 2008, Vol. 168) supports the idea that phosphate-based bowel cleansing regimens may result in worsening renal function. In a retrospective analysis looking at patients with normal creatinine levels at baseline who underwent colonoscopy with oral sodium phosphate solutions, they found that the GFR went slightly down compared to an age-matched control population. Prior to this, most data in support of phosphate-induced renal damage from colonoscopy preparations suggested that the only individuals at risk were those with pre-existing chronic kidney disease.

It will be interesting to see whether or not this will change clinical practice within the field of gastroenterology; for nephrologists it highlights the importance of asking patients about recent colonoscopies as part of the history-taking in patients experiencing acute kidney injury.

PS: 2 landmarks for the "Renal Fellow Network" today: (1) this is post #100, and (2) we are now the #1 "google hit" when the term "renal fellow" is searched...

The Latest on FGF23

There has been a lot of buzz surrounding the recent NEJM article by Gutiérrez (a recently graduated fellow from my program!) et al regarding the potential role of measuring FGF23 levels as an important biomarker in patients with ESRD.

FGF23, as detailed elsewhere in this blog, is felt to be the main phosphaturic hormone which regulates in vivo phosphorus balance. In a nest case-control sample consisting of 200 dialysis patients who died within the first year of initating dialysis and 200 dialysis patients who survived their first year of dialysis, it was determined that FGF23 levels were markedly elevated in the group who died within one year.

Obviously, this remains an association and does not imply causality. Indeed, one might postulate that the higher FGF23 group was sicker because they waited until uremic to begin dialysis. However, FGF23 remained associated with a higher mortality even when adjusted for phosphorus level, and the authors suggest that FGF23 may be a biomarker to determine which patients with CKD may be at risk for a high cardiovascular mortality and might benefit from early treatment with phosphorus binders.

FGF23

For many years, the existence of "phosphatonins"--substances secreted by certain tumors which result in profound renal phosphorus wasting and resultant osteomalacia--has been postulated.

Evidence has recently been accumulating that fibroblast growth factor 23 (FGF-23) is the phosphatonin we have been searching for:

High circulating levels of FGF23 are associated with hypophosphatemia, decreased 1,25 (OH) vitamin D levels, and rickets/bone disease.

The disease autosomal dominant hypophosphatemic rickets is caused by gain-of-function mutations in FGF23caused by splice site mutations. Conversely, the genetic disease inherited tumoral calcinosis, characterized by hyperphosphatemia, increased 1,25 (OH) vitamin D levels, and metastatic calcifications.

In ESRD, FGF23 levels are appropriately elevated in response to hyperphosphatemia, but due to a reduced GFR is unable to induce adequate phosphaturia.

How it works at a molecular level: FGF23 interacts with FGF receptors at the proximal tubule, resulting in decreased Na-PO4 exchange, as well as decreasing 1,25-alpha hydroxylase activity.