Potassium disorders

evaluation of hyperkalemia

Hyperkalemia is one of those disorders you should be able to treat with spinal-level reflexes. These reflexes are built with a prior knowledge of the different situations in which hyperkalemia can occur. Hyperkalemia will happen in your patients and so it’s best to immediately manage it and go forward with addressing all of the other patient issues for the day.

When faced with hyperkelamia, evaluate the severity first. Mild hyperkelamia is 5.1-5.5 mmol/L; moderate hyperkalemia is 5.6-6.0 mmol/L. Severe hyperkalemia is >6.0 mmol/L and most physicians would give rapid-acting medications for any potassium >6.5 mmol/L. After this point, decide if the high potassium value is actually real. Pseudohyperkelamia is any high potassium value which is not reflective of the potassium level throughout the patient’s circulation.

Notice that hyperglycemia was not listed in the table. The reason for that is because hyperkelamia due to hyperglycemia is true hyperkelamia. Hyperosmolality creases solvent drag that moves potassium out of the cells. In addition, you have a deficiency of insulin if your glucose level is extremely high. As insulin is one thing that moves potassium into cells, hyperglycemia by default means that insulin-mediated potassium shift into the cells is compromised. The good thing about a potassium level of 6.5 mmol/L in a patient with a blood sugar of 500 mg/dL is that you just give insulin and the potassium level takes care of itself.

If by this time, you have determined that the patient has true hyperkalemia which is not just due to hyperglycemia, the next step in evaluation is to estimate the chronicity of the hyperkalemia. 98% of total body potassium stores is intracellular. We have 70 mEq potassium in our extracellular stores and 3500 mEq in our intracellular stores. This means that, in a young 70kg man with a serum potassium level of 5.0 mmol/L, a 0.8% shift of the intracellular potassium stores into extracellular space will cause the serum potassium to rise to 7.0 mmol/L. It’s as if our serum potassium levels are standing at the base of the Hoover Dam. Any slight leak in the dam, which leaks intracellular potassium can be lethal. If there is an acute rise in serum potassium, then the hyperkelamia is likely due to a trans-cellular shift. If there is persistent hyperkalemia, then it is almost certainly due to a defect in renal excretion. We eat potassium and excrete it for the kidneys. There are two parts to this equation, but the rate-limiting step is always due to a defect in renal excretion of potassium. To drive home this point, lets point out what happens if people ingest insane amounts of potassium. In one study (Rabelink 1990), healthy adults increased potassium intake from a baseline of 100mEq daily to a very high amount of 400mEq daily for 20 days. The baseline K+ was 3.8 mmol/L. At the end of day 2 of the very, very, very, very high potassium diet, the serum potassium was only 4.8 mmol/L. By day 20, the serum potassium was only 4.2 mmol/L. The differential diagnosis of chronic hyperkalemia is one thing — reduced renal potassium excretion. There are multiple medications and renal conditions that can cause reduced renal excretion. Find a list of these and check your patient for these conditions.

Treatment of Hyperkalemia

There are two types of true hyperkalemia — hyperkelamiea without associated ECG changes or hyperkalemia with ECG changes. Peaked T waves are the most commonly cited hyperkalemia-associted ECG changes, but there are a number of different manifestations. Firstly, there is no set potassium level at which ECG changes occur, but when they do, they indicate myocardial instability which must be immediately addressed.

ECG changes that occur in association of hypernatremia are peaked T waves, flattened P waves, prolonged PR interval, and widened QRS. If any of these changes are present in a patient with hyperkalemia, first give 1 gram calcium gluconate. This medication can be given trough a peripheral line. Calcium chloride cannot be given through a peripheral line as it carries the risk of tissue necrosis. It begins immediately through direct chemical antagonism. Next, give 10 units IV insulin (not subcutaneous), a 25 gram push of D50, and start a D5W drip a 75mL/hr to prevent hypoglycemia. If the patient has blood glucose >300mg/dL, don’t give any D50 or a D5W drip. Also give 10-20mg inhaled albuterol. Remember that this is 4-8x the normal albuterol dose and so the respiratory therapist may come asking for an explanation before they know why you are giving it. Over the course of an hour, you can lower serum K+ by 1.2 mmol/L. The combination of all of this treatment stabilizes the myocardium immediately, begins to shift K+ into cells within 30 minutes and this effect lasts for 4 hours.

The next step is to remove potassium out of the body. It is likely that most patients will have some degree of volume overload and so loop diuretics will be useful. If severe hyperkalemia is present, the patient should get a foley to monitor urine output. If the patient has oliguric AKI, give lasix 80-120mg IV. If they aren’t producing at least 100mL urine per hour by one hour after the first dose, double the dose to 160mg-240mg IV. If that doesn’t work, then dialysis is likely needed for potassium removal if indicated by the particular situation. If the patient has only CKD or non-oliguric AKI with slowly rising creatinine, then multiply their serum Cr by 20 and give that much lasix. So, if the Cr is 2.0 — give 40mg IV lasix IV BID and readjust the dose every 12 hours. If the Cr is 4.0, start with 80mg IV BID. Lasix doesn't last forever and so you can always give IV fluid boluses or NS 500mL/hr if you overshoot your intended urine output. At this point, some questions may arise… what if the patient is volume depleted? In this case, it is possible that prerenal azotemia from true volume depletion is present. Give normal saline if the serum bicarb level is normal. If the serum bicarb level is low and it’s due to metabolic acidosis, then add sodium bicarbonate into appropriate IV fluids to make an isotonic crystalloid solution (75mEq sodium bicarbonate in 1/2 NS. If the patient is very acidotic, then you put 150mEq sodium bicarbonate into 1L D5W). When the prerenal portion of renal failure corrects, then it is possible that the patient will excrete potassium. If the patient is evolvemic, then it’s perfectly alright to give normal saline and loop diuretics since our goal in this situation is not to change volume status, but to encourage tubular flow of urine to encourage potassium excretion in the distal tubule.

If you need to use the GI tract to get potassium out of the body, zirconium cyclosilicate is currently the best treatment. Kayexalte carries a risk of colonic necrosis that is estimated to be 0.27% to 1.8% though, and so keep this in mind. It should not be given in any patients with ileus, suspected ileus, post-operative patients, or any patient in whom you suspect that there is any altered bowel function. Veltassa is not appropriate for the inpatient setting as it takes too long to work.

Lastly, dialysis is the ultimate way to get potassium out of the body. A 2h dialysis session will lower the potassium. If a patient with severe hyperkalemia fails medical management of hyperkallemia, then dialysis is likely indicated.

Evaluation of Hypokalemia

In contrast to hypERkalamia, hypOkalemia has a larger differential and nuanced diagnosis. When that serum potassium comes back low as a red number instead of a black number, the first step is to decide if you are actually viewing true hypokalemia. Metabolically-active cells such as those seen in acute myeloid leukemia can cause pseudohypokalemia. In this case, once blood is in the test tube, the leukemia cells take up potassium thus making serum potassium look artificially low. Thrombocytosis would not cause the same issue. Pseudohypokalemia can be prevented by rapidly separating plasma and cells or storing the blood at 4 degrees C.

If there is true hypokalemia, then the following is the next step and involves an evaluation of the differential diagnosis for transcellular shifts. We briefly talked about the mechanism for hypERkalemia resulting from acidosis. Either metabolic or respiratory alkalosis can promote potassium entry into cells. In the presence of alkalemia, hydrogen ions leave the cells to minimize the increase in extracellular pH; the necessity to maintain electroneutrality then requires the entry of some potassium (and sodium) into the cells. The serum potassium concentration falls by less than 0.4 mEq/L for every 0.1 unit rise in pH. Beta-2 adrenergic agonists as well as insulin increase activity of Na-K-ATPase and thus cause potassium shift into cells. Barium blocks K+ channels, thus inhibiting exit of potassium, brought in by Na-K-ATPase (by the way, if you ever see a case of this, write a case report on it). Hypokalemic periodic paralysis has two main types. The inherited form is due to an alteration in the alpha-1 subunit of the voltage-gated calcium channel. The second type is an acquired form associated with hyperthyroidism and is termed thyroxotic hypokalemic periodic paralysis. In this disorder, there is a defect in potassium channels in cells. In this context, the effect of anything that shifts potassium into cells is amplified. The classic case is someone with the sudden onset of weakness after a large carbohydrate meal. This happens because insulin causes an increase in the activity of Na-K-ATPase pumps which can then not escape.

If pseudohypokalemia and transcellular shifts are not the cause, then the other main cause would be an actual loss of potassium from the body. This is either due to renal potassium wasting or GI loss of potassium. Evaluation of renal potassium wasting is useful in this setting. Renal potassium wasting is present when 24h urine K+ secretion (on 24h urine collection) of >30mEq in a day or a spot urine creatinine-to-potassium ratio greater than 13mEq/g or a spot urine potassium >30-40. Basing renal potassium wasting based only on a spot urine potassium and not event a urine creatinine-to-potassium ratio is not great, but seeing a high urine potassium in the setting of hypokalemia is a great clue which can prompt further testing. The spot urine potassium would not be useful if the urine osmolality is lower than the plasma osmolality or if the urine sodium is lower than 30mEq/L.

If urine studies suggest renal potassium wasting, than make sure you rule out hypomagnesemia. Next, evaluation of alkalosis, aldosterone, cortisol and blood pressure should take place. The first division takes place if the blood pressure is high or low and basically distinguishes Bartter and Gitlemans from other disorders. Patients with these two conditions will not be hypertensive (We assume that we made sure that the patient is not on a loop or thiazide diuretic). The classic combination of hypertension, hypokalemia, metabolic alkalosis along with high aldosterone and suppressed renin is due to the umbrella term primary aldosteronism. This umbrella term includes aldosterone producing adenomas (65%) as well as bilateral adrenal hyperplasia, GRA, unilateral adrenal hyperplasia, aldosterone-producing adrenocortical carcinomas and ectopic aldosterone-producing tumors. GRA is caused by an unequal crossover between chromosomes 16 and 18 which places the aldosterone synthase gene under control of the promoter for cortisol. Therefore, aldosterone production is inhibited by glucocorticoids. High renin and high aldosterone suggests secondary aldosteronism and includes Bartter syndrome, Gitleman syndrome, RAS, poorly reabsorbed anions, and loss of gastric secretions. Bartter’s syndrome has 5 different types, but all types essentially create the same effect as a loop diuretic. If you can even arrive at a diagnosis of Bartter syndrome, then congratulations, you are a master clinician. Gitleman syndrome takes place in the DCT and is a defect in the NCC channel. It acts like a thiazide diuretic. The lumen-negative electrical gradient created by sodium reabsorption in the cortical collecting tubule is partially attenuated by chloride reabsorption. There are, however, a number of clinical settings in which sodium is presented to the distal nephron with relatively large quantities of a nonreabsorbable anion, including bicarbonate with vomiting or proximal renal tubular acidosis, beta-hydroxybutyrate in diabetic ketoacidosis, hippurate following toluene use (glue sniffing), or a penicillin derivative in patients receiving high-dose penicillin therapy (or Zosyn!). In these settings, more of the delivered sodium will be reabsorbed in exchange for potassium, leading to a potentially marked increase in potassium excretion. The syndrome of apparent mineralocorticoid excess is due to inhibition of 11-beta-hydroxysteroid dehydrogenase type 2 which causes leaves cortisol free to acitivate the MR. Cushing syndrome accomplishes the same goal as the syndrome of apparent mineralocorticoid excess, but does so by flooding this area with more cortisol that 11-beta-hydroxysteroid dehydrogenase type 2. Diagnosis involves increased ratio of free urinary cortisol to cortisone. Treatment is with either ramiloride or spironolactone. Congenital adrenal hyperplasia (CAH) leaves processor molecules in excess which activate the MR. Essentially the only disorder that we need to know for the boards is CAH due to 11-beta-hydroxylase deficiency. This enzyme mediates the final step in cortisol synthesis. It causes excessive androgen production and can virilize a genetically female fetus. Diagnosis is confirmed by demonstration of marked elevations of 11-deoxycortisol and 11-deoxycorticosterone (DOC), the substrates of 11-beta-hydroxylase. Treatment is life-long glucocorticoids. Activating mutation of the mineralocorticoid receptor presents as early onset hypertension exacdrbated by pregnancy. Prednisone activates the MR. Sprionlactone paradoxically activates the MR and raises blood pressure.

REFERENCES

Rabelink, T. J., Koomans, H. A., Hené, R. J., & Mees, E. J. D. (1990). Early and late adjustment to potassium loading in humans. Kidney international, 38(5), 942-947.