The Role of the Gut for Potassium Homeostasis in CKD

Recently the FDA approved one of the newer exchange resins (Lokelma, ZS-9) for the treatment of hyperkalemia. (Veltassa, Patiromer), a calcium based cation exchange resin, was recently developed and approved since 2015 for the treatment of hyperkalemia. These two new drugs, which each have large randomized placebo controlled trials confirming their effectiveness, now seem bound to replace the ancient but much experienced Kayexalate (Sodium Polystyrene Sulfonate; SPS) for the management of hyperkalemia. However, they are not approved in Canada and their cost makes them less enticing options in a publicly funded health care system. Also, what about safety concerns? SPS has been used since the 1950s, however it took decades for case reports to come out possibly linking fatal colon necrosis to Kayexalate use. This association remains hotly debated. A systematic review in 2013 found a total of 58 cases of biopsy confirmed bowel injury that were “possibly” related to SPS use based on the WHO causality criteria. Interestingly, these cases were not limited to patients using SPS with sorbitol, nor to the post-operative population as was once thought to be the major risk factors for adverse GI events from SPS. On the other hand, a retrospective single-centre review over a 10 year period found only 3 cases of colon necrosis among 2,194 patients having received SPS. When they compared this incidence of colon necrosis to control patients having not received SPS, they found a non-significant RR of 2.10 (0.66-6.64; P=0.2). Many nephrologists will argue that SPS has a long standing proven track record and that colon necrosis is such a rare side effect that it is unlikely to be related. While no one knows if this is truly the case, the uncertainty surrounding this issue is one of the major reasons that led to the development of the newer exchange resins: ZS-9 and patiromer.

Before praising and widely accepting these newer, expensive products, would it not be wise to use the same caution we would for SPS? It can take years for such a rare adverse event as colon necrosis to manifest itself, how do we know we won’t be facing this GI safety issue 5 to 10 years from now?

Perhaps we could look for other ways of trying to enhance GI elimination of potassium.

--> BK-channel in colon enterocyte (Source: Sandle et al. QJM. 2010 Feb;103(2):85-9)    
The colonic enterocyte has an ability to excrete potassium in stool. This may be an important physiologic property in patients with advanced CKD who lose the ability to renally excrete potassium. In fact, patients with ESKD are known to have a greater excretion of potassium in their stool compared to individuals with normal kidney function. An important transporter involved in this process is the BK-channel located on the apical side of the colonic enterocyte within the colon crypts (see Figure). This channel is responsible for actively secreting potassium in the GI lumen. Aldosterone stimulates this channel, enhancing K secretion in the stool, similar to its effect in the kidneys. In fact, a study in the 1970s of patients suffering from acute cholera showed that within 12 hours of receiving a single dose of 100mg of spironolactone, there was a significant decrease in stool loss of potassium and an increase in stool loss of sodium. This likely is a major reason explaining why anuric patients with ESKD may still have from hyperkalemia if given RAAS blockade. The BK channel will also be stimulated by various diarrheal states such as Ogilvie’s syndrome, villous adenoma and certain laxatives such as bisacodyl. Bisacodyl is a stimulant laxative which enhances peristalsis through parasympathetic nerve activation, but also stimulates cAMP production within the colon enterocyte. This cAMP is thought to enhance K excretion via the BK-channel. A study performed in 2005 used immunohistologic analysis of colonic enterocytes in patients with ESKD and showed that BK-channels are upregulated in these patients compared to controls with normal kidney function. The upregulation of BK-channels probably represent an adaptation to chronic potassium load in order to maintain homeostasis. Interestingly, a small study in 2003 tested the effect of Bisacodyl on potassium in patients with ESKD. Eight controls normal kidney function and 13 patients with ESKD were given Bisacodyl 5-10mg PO titrated to achieve 2 soft bowel movements per day and these were compared to 5 ESKD patients given lactulose 10ml PO titrated also to 2 soft bowel movements per day. After 2 weeks of treatment, they found that the ESKD patients with Bisacodyl had a significant decrease in potassium from 5.9 to 5.4mmol/L whereas the controls given Bisacodyl and the ESKD patients given lactulose had no change in potassium values after 2 weeks of therapy. This suggests that Bisacodyl increases potassium excretion in the stool in patients with ESKD through a mechanism not simply related to its laxative effect, but likely through stimulation of the BK-channel of the colonic enterocyte. It is unfortunate that no further studies have looked into this therapeutic option for maintaining potassium homeostasis in patients with advanced CKD. Unfortunately, we don’t know if Bisacodyl would work as well in the non-dialysis population since we could expect these patients may not have as upregulated BK channels as in ESKD who are faced with more chronically elevated potassium loads. However, this would be an interesting study to pursue since Bisacodyl would be a cheap and safe way of controlling potassium in patients with advanced CKD by enhancing the body’s natural adaptation to potassium handling.

David Massicotte-Azarniouch
Nephrology Fellow
University of Ottawa

Novel hyperkalemia treatments – practical points

I was quite excited to see the development of novel agents to help us manage hyperkalemia in our kidney patients in place of kayexalate, which is extremely unpleasant for patients, is associated with diarrhea and have reported severe side effects such as intestinal necrosis. As previously reviewed, recent trials were published on two major drugs: Patiromer (Veltassa) and ZS9.

 Patiromer is a polymer that binds K in exchange for calcium (NEJM 2015). FDA recently approved this agent for treatment of chronic hyperkalemia (not acute) though it included a black box stating that patiromer can significantly affect the absorption of many orally administered medications and therefore it should be taken 6 hours apart from any other medication (a significant limitation). This decision by the FDA was based on some in vitro data (could not find the actual data on the recent papers) so it is unclear which drugs may significantly be affected if taken in close proximity. The company is now performing a number of studies to elucidate that. Currently, Patiromer is available to patients but it is only dispensed in subspecialty pharmacies and most hospitals don’t have this medication on formulary yet, in part related to the cost of $30 per pill. The main side effects reported with Patiromer were hypomagnesemia and constipation. Duration of trials were up to 52 weeks.

ZS9 is a crystalline compound of zirconium silicate that exchanges Na+/H+ for K+. Based on its mechanism of action, patients will have about 300mg of extra Na+ intake per day (Kayexalate has the same issue). The main side effect of ZS9 was edema (~7%). Duration of trial with ZS-9 was only 4 weeks. FDA has not approved ZS9 yet but it is likely that once approved, it will be the leading drug in the market, especially if the black box from FDA remains for patiromer.

 These first trials excluded patients on dialysis and patients with potassium greater than 6.5 mmol/L. Let us know if you have used Patiromer in the real world so that we can start gaining more knowledge outside of a controlled clinical trial.

Figure from Nephcure.org

Hyperkalaemia, ICDs and ECG changes.

I recently admitted a woman overnight with prerenal AKI and serum K+ of 7 mmol/L. A passing ED physician mumbled to me not to worry as their ECG didn’t require calcium gluconate, and besides, they had an ICD inserted recently.

As I looked over the ECG, I realised I had no idea what effect hyperkalaemia would have on a pacemaker. Furthermore I always suspected that ECG changes were not a sensitive method of alerting physicians to danger and just delayed treatment. So firstly, what effect does hyperkalaemia have on those pacing/ICD leads?

A recent review from europace describes 3 key changes:

1. The QRS complex can widen due to intraventricular conduction delay

2. The increased ventricular pacing thresholds can cause failure to capture

3. It can lead to over-sensing of spontaneous T waves and inappropriate shocks if the device is an ICD

An hour later this lady had multiple long runs of ventricular tachycardia and lost consciousness .The overdrive pacing spikes could be seen on the monitor but there was no electrical capture and after 20 seconds the device ultimately defibrillated the patient each time. Could this have been avoided with quicker initial treatment and administration of calcium? 

Given the potentially disastrous consequences of hyperkalaemia in a patient with a pacemaker, just fix it, and don’t rely on the ECG to reassure you!

So what about my suspicion that ECG changes are not particularly sensitive?

It would seem that physicians cannot predict clinical severity based on ECG changes alone. A blinded reading of 200 ECGs gave a best  positive predictive value for the presence of hyperkalaemia of 0.65. In addition to this, another study  in CJASN could find no correlation between T waves changes and serum K+. The problem worsens in dialysis patients who appear even less likely to ever manifest ECG changes. This is associated with having a higher serum calcium. Furthermore there are multiple case reports of profound hyperkalaemia both in patients with and without CKD who do not manifest any ECG changes. It is hypothesised that perhaps it if the rate of change that produces the ECG abnormalities, which makes some sense as to why the RRT cohort do not manifest these. 

Furthermore it has been shown worsening stages of CKD tolerate any given level of hyperkalaemia better, again perhaps due to the rate of change. I could find no good study showing that ECG changes, except for sinusoidal waves, predicted death or morbidity. The long described progression of t wave changes through to QRS widening and then sinusoidal waves is a physiologically pleasing one, it feels right and is very logical. The ECG itself however, would seem not to be particularly useful in treating real patients, and I am going to force myself to ignore normal ECGs. I think my assessment of the urgency will centre around how the patient looks and the rate of flux, background CKD, serum K+, likelihood of rapid improvement, medications and cardiac history.

A final point I stumbled across which I found interesting was the interaction between verapamil and hyperkalaemia. Both dogs and humans on verapamil have profoundly lower cardiac indexes when hyperkalaemic. I’m not sure is it the Ca2+ blocking effect worsening the hyperkalaemic effect on myocardium or the cumulative negative inotropy, but I can see a logical argument for empirical calcium therapy in those on verapamil and perhaps other calcium channel blockers.
Any thoughts?

A cause and a cure of hyperkalemia? The next #NephJC

There has been a flurry of publications in the field of hyperkalemia with 3 separate trials of oral potassium binding agents within a week of each other (Sodium Zirconium in JAMA, and NEJM and Patiromer in NEJM) and a potentially related observational trial on the risks of co-trimoxazole in patients on RAAS blockade in the BMJ. With all that reading to get through, the next NephJC on Tuesday Dec 2nd will be a double whammy. We will look at the HARMONISE trial of ZS-9, and a large study of co-trimoxazole and potential associations. 

Trim-Sulfa and Sudden Death in patients receiving inhibitors of renin-angiotensin system. 

The first paper for discussion is a large, Canadian, case control series, by the Canadian Drug Safety and Research Effectiveness Network, published in the BMJ. The hypothesis is the risk of sudden death in patients on RAAS blockade is higher following administration of specific antibiotics rather than amoxicillin. To answer their question, they searched 17 years of records representing over 1.6 million patients. They identified 39,879 with a label of sudden death and a subsequent group of 1,027 that had a prescription for the target antibiotics in the 7 days prior to dying. 
The authors write: “In the primary analysis, co-trimoxazole was associated with a significantly increased risk of sudden death within seven days relative to amoxicillin (OR 1.8 C.I 1.5-2.24)” Ciprofloxacin was associated with a somewhat lower risk of sudden death. I found it strange that norfloxacin, which has similar QT prolonging properties to ciprofloxacin, had had no such risk. The authors speculate this observed association may be due to trimethoprim’s activity as an ENaC antagonist. There are a number of important limitations to consider. There was no indication for antibiotics recorded. Also, the cases and controls had some important differences in terms of diuretic use and co-morbidities. Only 8.2% of the cases had renal disease, the stage of which was unclassified. The authors can only speculate about a possible mechanism involving hyperkalaemia as no K levels were obtained for any of these patients, nor any ECG to help explain the effect of ciprofloxacin. 

Harmonise: Effect of Sodium Zirconium Cyclosilicate on Potassium Lowering for 28 Days Among Outpatients With Hyperkalemia. 

ZS-9 is a zirconium silicate, a non-absorbable potassium binding agent. It is an inorganic cation exchanger crystalline with the capacity to bind both potassium and ammonium in the GI tract. Its creators tout its non-absorbable nature as the key to minimising systemic side effects. HARMONISE is a phase 3, multicenter, randomized, double-blind, placebo-controlled trial spanning 44 centres. Inclusion criteria was simply a serum K of ≥ 5.1 on 2 occasions. Initially, 258 patients who met eligibility criteria were given ZS-9 10g three times daily. If they achieve normokalaemia within 48 hours, they were then randomized to a placebo, or increasing doses of ZS-9 once daily. The mean eGFR was 46 ml/min/1.73m2 and no ESKD patients are represented. 
Did it work? The short answer is yes. ZS-9 had a reasonable rapid rate of onset and within 2 hours, serum Potassium has dropped by −0.4 mEq/L (95% CI, −0.5 to−0.4) and was - 1.1 mEq/L by 48hours. Encouragingly, it seems generally well tolerated with some edema and hypokalemia as the doses increased. In conclusion, this is a well executed phase 3 trial and ZS-9 has potential to be a well tolerated and predictable treatment option for hyperkalemia. The authors quite rightly point out we still have no data beyond 4 weeks, nor have we any meaningful endpoint such as mortality or hospital admissions. It is an encouraging study none the less, and should lead to FDA approval and another tool in our kit.

Full post can be seen at www.nephjc.com 

Authored by Eoin O'Sullivan.

Concept Map of Hyperkalemia

Pseudohyperkalaemia with extreme leucocytosis

We were following a CKD patient with CLL in hospital with the haematology team. One morning, out of the blue, his potassium came back at 6.5mmol/L. It had been stable in the high four range for the duration of his admission to that point. There were no changes to his diet or medications – nothing to blame that we could ascertain.

When the team called us, we suggested repeating the lab draw stat, taking care to minimize any chance of haemolysis (avoid tourniquet etc.), and to perform an ECG while we were waiting for the repeat result to come through. The ECG showed normal sinus rhythm and no changes associated with hyperkalaemia; the patient looked well, but the repeat serum sample came back at 6.7mmol/L!!!

Very importantly, looking through the rest of his lab results we noted his peripheral white cell count was shooting up (>100K). A spark of inspiration from one of the attendings wondered if this could be pseudohyperkalaemia from extreme leucocytosis?

Potential mechanisms for pseudohyperkalaemia in cases of leucocytosis and thrombocytosis include:

1. release of K during the clotting process of collected blood 
2. abnormally high fragility of the leucocytes. 

Other factors that increase tendency for lysis include increased temperature, length of sample standing time, use of pneumatic blood drawing devices, tourniquets and ‘difficult sticks’. Due to the influence of the clotting process, plasma samples more closely reflect the true potassium level than serum samples in these patients.

If a case of pseudohyperkalamia is suspected, then it’s important to figure out what type of tube the blood is collected in at your hospital. Simultaneous samples should be taken for plasma and serum processing. If pseduohyperkalemia is present, the plasma sample will be the more accurate – we sent a sample in arterial blood gas syringe (which are heparinized, processed very quickly and the results based on plasma samples) – the K came back at 4.8mmol/L, avoiding unnecessary and potentially dangerous treatment.

A final point in ensuring patient safety in this scenario is accurate handover to night-staff or change of shifts – make sure everyone knows to order and treat according to the plasma sample.

See Nate's post for other causes of pseudohyperkalaemia or Justin's post on a similar patient presentation.

What is your K+ threshold for dialysis?

Hyperkalemia can be a life-threatening condition. Nonetheless, in transplant recipients with delayed graft function, there is a tendency to try to avoid dialysis early after transplant. The reason behind this approach is that a session of hemodialysis further delays the recovery from ATN/ischemia and significantly drops urine output afterwards (indirect evidence - no randomized trials available). This is potentially related to hemodynamic changes associated with the dialytic procedure, even in the absence of volume removal (osmotic shift). So what are your options if you have a patient with a potassium of 6 about 72 hours after transplantation with low urine output?

First option, if he doesn’t have ECG changes, is to try a loop diuretic at a high dose to see if he can respond to that and produce some kaliuresis. If he doesn’t respond and has had a bowel movement since surgery, I would do a trial of sodium polystyrene sulfonate binding resins (e.g.Kayexalate). Now, it gets into a little controversial territory.

Last year, Sterns et al. in an editorial on JASN concluded that the use of SPS resins is largely unproven and potentially harmful, and should be considered only as a last resource. That was quite shocking to me since I have used it many times and I always saw a decrease of serum potassium levels. I do acknowledge the increased risk of colonic necrosis that has been reported with sorbitol commonly added to Kayexalate to increased bowel movements, but I was prescribing only mixed with water and other than the low palatability of the solution, no major side effects were seen.

Actually, after reviewing the literature, the most common side effects reported are nausea, vomiting and constipation. I was able to download the first trial using SPS resins published in the NEJM in 1961! Small number of patients (n=22) but the results looked quite impressive, since a median dose of 40g was able to decrease the potassium level by 1mEq/L and additional doses led to further decrease in levels (median decrease of 1.8mEq/L). It is important to know that the peak effect of SPS takes about 4-6 hours, so if ECG changes are present, you will need to use IV calcium for membrane protection and albuterol/insulin+dextrose for transient K+ shift to the intracellular. The majority of reported complications were associated with SPS/70% sorbitol enema. This combination should be avoided, especially in patients with compromised GI function (e.g. ileus post-op). As an alternative, lactulose could be used to stimulate the GI tract. Finally, Kayexalate releases sodium ions after uptake of potassium, so there is a potential side effect of edema due to sodium retention (not significant in 1-2 doses).

Is there a way to estimate how much resin you will require? Efficiency of a resin is dependent on several factors, including intrinsic properties of the resin (capacity and ion selectivity), [K+] and [Na+] extracellular concentrations and colonic transit time. In general, a decrease in 0.5mEq/l [K+} would require at least 30g of Kayexalate.

Going back to our patient: After a diuretic trial, I would recommend two doses of 30g of Kayexalate mixed with water PO with repeat labs in 8-12 hours. Caveat, if the patient has no urine output or have other potential indications for dialysis (e.g.volume), you should just buy the bullet and dialyse him.

More acute hyperkalemia management on this blog. Picture above from Jackson Hole - insane ski destination...