Extracellular potassium concentration is normally maintained between
4.0 and 4.5 mEq/L by a complex interplay of potassium excretion and
consumption. Ninety-five percent of total body potassium is intracellular;
only 2% is extracellular. A 70-kg man, for instance, has about 3,920 mEq
of potassium in the intracellular space but only 59 mEq in the
extracellular space.1 Given that the total daily intake of
potassium from a normal diet can be up to 200 mEq, one can see how
precisely and quickly the body must be able to respond to any given
potassium load in order to prevent severe hyperkalemia.
Total body potassium levels are regulated mostly by the kidneys, with
only 5% to 10% of ingested potassium excreted in the feces.1
Renal excretion of potassium is determined by the rate of potassium
filtration across the glomerular basement membrane and by the rate of its
secretion and resorption in the distal tubules of the nephron. When
increased intake of potassium overwhelms the ability of the kidneys to
excrete potassium, or when a decrease in renal function occurs,
hyperkalemia may result. Because there are often no clinical signs or
symptoms to suggest hyperkalemia, clinicians must frequently rely on
clinical information (that is, a history of renal failure or the ingestion
of medications known to cause hyperkalemia), laboratory data, and
electrocardiographic changes to make the diagnosis.
Hyperkalemia is a common cause of the cardiac arrhythmias seen in
clinical practice. The challenge in managing hyperkalemia comes from the
fact that it can be difficult, if not impossible, to identify the
condition solely on the basis of electrocardiographic information.
Patients who present with hyperkalemia may have a normal electrocardiogram
or have changes that are so subtle that physicians frequently have
difficul-ty attributing these changes to increased potassium levels. In a
study performed at the University of Pittsburgh Medical Center, only 46%
of patients with potassium levels greater than 6.0 mEq/L had
electro-cardiographic changes, and only 55% of patients with potassium
levels greater than 6.8 mEq/L had changes consistent with
hyperkalemia.2 In fact, there have been several reports in the
literature of patients who had potassium levels greater than 7.5 mEq/L
with no electrocardiographic manifestations of hyperkalemia.3
--5 Even when there is evidence of hyperkalemia on a patient's
electrocardiogram, physicians often miss the diagnosis. Wrenn and
colleagues6 designed a study to determine the ability of
physicians to predict the presence of hyperkalemia solely on the basis of
their patients' electrocardiograms. The physicians in this study were able
to predict hyperkalemia with a sensitivity of 35% to 43% and a specificity
of 85% to 86%. This small study further emphasizes how difficult
hyperkalemia can be to diagnose. Nevertheless, hyperkalemia can manifest
with classic electrocardiographic changes that suggest its presence.
Effects of Hyperkalemia on Impulse
Production and Propagation
Potassium and sodium concentrations in the
intracellular and extracellular compartments play a vital role in the
electrophysiologic function of the myocardium. Concentration gradients are
established across the myocyte membrane secondary to very high
intracellular potassium concentrations and a relative paucity of potassium
ions in the extracellular space. The opposite is true of sodium ions,
which are abundant extracellularly and relatively few intracellularly.
These concentration gradients are maintained by sodium-potassium adenosine
triphosphatase (Na-K ATPase) pumps on the cellular wall, which actively
pump sodium out of the myocyte and potassium inward. These concentration
gradients establish an electrical potential across the cell membrane,
leading to a resting membrane potential of [minus sign]90mV. The potassium
gradient across the cellular membrane is the most important factor in
establishing this membrane potential; therefore, any changes in
extracellular potassium concentration may have profound effects upon
myocyte electrophysiologic function.7 For instance, as
potassium levels increase in the extracellular space, the magnitude of the
concentration gradient for potassium across the myocyte diminishes, thus
decreasing the resting membrane potential (that is, --90 mV to --80 mV;
see Fig. 3).
Phase 0 of the action potential occurs when voltage-gated sodium
channels open and sodium enters the myocyte down its electrochemical
gradient (Fig. 3). The rate of rise of phase 0 of the action potential
(Vmax) is directly proportional to the value of the resting
membrane potential at the onset of phase 0.7 --9 This is
because the membrane potential at the onset of depolarization determines
the number of sodium channels activated during depolarization, which in
turn determines the magnitude of the inward sodium current and the
Vmax of the action potential. As illustrated in Figure 4,
Vmax is greatest when the resting membrane potential at the
onset of the action potential is approximately [minus sign]75 mV, and does
not increase as the membrane potential becomes more negative. Conversely,
as the resting membrane potential becomes less negative (that is, [minus
sign]70 mV), as in the setting of hyperkalemia (Fig. 3), the percentage of
available sodium channels decreases. This decrease leads to a decrement in
the inward sodium current and a concurrent decrease in the
Vmax; therefore, as the resting membrane potential becomes less
negative in hyperkalemia, Vmax decreases. This decrease in
Vmax causes a slow-ing of impulse conduction through the
myocardium and a prolongation of membrane depolarization; as a result, the
QRS duration is prolonged.
As previously discussed, increasing the extracellular potassium
concentration results in a decrease in the resting membrane potential
(that is, from [minus sign]90 mV to [minus sign]80 mV). In turn, the
threshold potential decreases (that is, from [minus sign]75 mV to [minus
sign]70 mV); this 5-mV decrease, however, is less than the decrease in
resting potential. Therefore, the difference between the resting and
threshold potentials decreases to approximately 10 mV (as opposed to 15 mV
in a physiologic milieu). As potassium levels increase further, the
resting membrane potential continues to become less negative, and thus
progressively decreases Vmax. The changes in threshold
potential now parallel the changes in resting potential, and the
difference between the two reaches a constant value of approximately 15
mV. The decrease in Vmax levels causes a slowing of myocardial
conduction, manifested by progressive prolongation of the P wave, PR
interval, and QRS complex. In summary, the early effect of mild
hyperkalemia on myocyte function is to increase myocyte excitability by
shifting the resting membrane potential to a less negative value and thus
closer to threshold potential; but as potassium levels continue to rise,
myocyte depression occurs and Vmax continues to decrease.
Hyperkalemia also has profound effects upon phase 2 and phase 3 of the
action potential. After the rapid influx of sodium across the cell
membrane in phase 0, potassium ions leave the cell along its
electrochemical gradient, which is reflected in phase 1 of the action
potential. As the membrane potential reaches [minus sign]40 to [minus
sign]45 mV during phase 0, calcium channels are stimulated, allowing
calcium to enter the myocyte. The maximum conductance of these channels
occurs approximately 50 msec after the initiation of phase 0 and is
reflected in phase 2 of the action potential.7 During phase 2,
potassium efflux and calcium in-flux offset one another so that the
electrical charge across the cell membrane remains the same, and the
so-called plateau phase of the action potential is created (Fig. 3).
During phase 3, the calcium channels close, while the potassium channels
continue to conduct potassium out of the cell; in this way, the
electronegative membrane potential is restored.7 One of the
potassium currents (Ikr), located on the myocyte cell membrane, is mostly
responsible for the potassium efflux seen during phases 2 and 3 of the
cardiac action potential.10 For reasons that are not well
understood, these Ikr currents are sensitive to extracellular potassium
levels, and as the potassium levels increase in the extracellular space,
potassium conductance through these currents is increased so that more
potassium leaves the myocyte in any given time period.10 This
leads to an increase in the slope of phases 2 and 3 of the action
potential in patients with hyperkalemia and therefore, to a shortening of
the repolarization time. This is thought to be the mechanism responsible
for some of the early electrocardiographic manifestations of hyperkalemia,
such as ST-T segment depression, peaked T waves, and Q-T interval
shortening.11,12
Surface Electrocardiogram Manifestations
of Hyperkalemia
In experimental models, there is a very orderly
progression of electrocardiographic changes induced by
hyperkalemia.13,14 The earliest electrocardiographic
manifestation of hyperkalemia is the appearance of narrow-based, peaked T
waves. These T waves are of relatively short duration, approximately 150
to 250 msec, which helps distinguish them from the broad-based T waves
typically seen in patients with myocardial infarction or intracerebral
accidents.7 Peaked T waves are usually seen at potassium
concentrations greater than 5.5 mEq/L and are best seen in leads II, III,
and V2 through V4, but are present in only 22% of
patients with hyperkalemia.8,11,15 It may be that increased
myocyte excitability, shortening of the myocyte action potential, and an
increase in the slope of phase 2 and 3 of the action potential account for
the T wave peaking seen in mild hyperkalemia.11
As serum potassium levels increase to greater than 6.5 mEq/L, the rate
of phase 0 of the action potential decreases, leading to a longer action
potential and, in turn, a widened QRS complex and prolonged PR interval.
Electrophysiologically, this appears as delayed intraventricular and
atrioventricular conduction.7,11 As the intraventricular
conduction delay worsens, the QRS complex may take on the appearance of a
left or right bundle branch block configuration. A clue that these
electrocardiographic changes are due to hyperkalemia, and not to bundle
branch disease, is that in hyperkalemia the conduction delay persists
throughout the QRS complex and not just in the initial or terminal
portions, as seen in left and right bundle branch block,
respectively.11,16 As potassium levels reach 8 to 9 mEq/L,
sinoatrial (SA) node activity may stimulate the ventricles without
evidence of atrial activity, producing a sinoventricular rhythm. This
occurs because the SA node is less susceptible to the effects of
hyperkalemia and can continue to stimulate the ventricles without evidence
of atrial electrical activity.11,17 The electrocardiographic
manifestations of continued SA node function in the absence of atrial
activity may be very similar to those of ventricular tachycardia, given
the absence of P waves and a widened QRS complex (Fig 1).
As the hyperkalemia worsens and the potassium levels reach 10 mEq/L,
sinoatrial conduction no longer occurs, and passive junctional pacemakers
take over the electrical stimulation of the myocardium (accelerated
junctional rhythm).11,12,18 If hyperkalemia continues unabated,
the QRS complex continues to widen and eventually blends with the T wave,
producing the classic sine-wave electrocardiogram. Once this occurs,
ventricular fibrillation and asystole are imminent.
In addition to the previously mentioned arrhythmias, many other
electrocardiographic abnormalities have been associated with hyperkalemia.
In patients with acutely elevated serum potassium levels, a
pseudomyocardial infarction pattern has been reported to appear as massive
ST-T segment elevation develops secondary to derangements in myocyte
repolarization.19 --23 Early stages of hyperkalemia may
manifest with only shortening of the PR and QT interval.8 Sinus
tachycardia and bradycardia, idioventricular rhythm, and 1st-, 2nd-, and
3rd-degree heart block have all been described on the presenting
electrocardiograms of patients with hyperkalemia.7 Given the
vast array of electrocardiographic manifestations of hyperkalemia, the
difficulty in consistently identifying hyperkalemia on the basis of
electrocardiographic abnormalities, and the fact that the
electrocardiogram during hyperkalemia may progress from normal to that of
ventricular tachycardia and asystole precipitously, physicians need to
consider this diagnosis in patients at risk.8
Causes of Hyperkalemia
Numerous
causes of hyperkalemia are seen in clinical practice. The most common are
renal disease and the ingestion of medications that predispose the patient
to hyperkalemia.2 Medications known to cause hyperkalemia
include angiotensin-converting enzyme inhibitors, angiotensin-receptor
blockers, penicillin G, trimethoprim, spironolactone, succinylcholine,
alternative medicines, and heparin, to name just a few.24 --30
In their study in a university setting, Acker and colleagues2
reported that 75% of all patients with severe hyperkalemia had renal
failure, and 67% were taking a drug that predisposed them to hyperkalemia.
Other less common causes of hyperkalemia include massive crushing injury
with resultant muscle damage, large burns, high-volume blood transfusions,
human immunodeficiency virus infection, and tumor lysis syndrome.8,31
--35 In many patients, the cause of hyperka-lemia is multifactorial
and never clearly defined.
Treatment of Hyperkalemia
Although
hyperkalemia is one of the deadliest electrolyte abnormalities, it is also
one of the most treatable. As previously discussed, the diagnosis of
hyperkalemia can be difficult if one relies solely on electrocardiographic
criteria. Frequently, physicians must initiate treatment for hyperkalemia
on the basis of a patient's clinical scenario (such as a cardiac arrest
occurring in a chronic dialysis patient). More commonly, however, the
patient is treated when laboratory data become available. Most authorities
recommend treatment for hyperkalemia when electrocardiographic changes are
present or when serum potassium levels are greater than 6.5 mEq/L,
regardless of the electrocardiogram.36,37 The treatment for
hyperkalemia can be thought of in 3 distinct steps. First, antagonize the
effects of hyperkalemia at the cellular level (membrane stabilization).
Second, decrease serum potassium levels by promoting the influx of
potassium into cells throughout the body. Third, remove potassium from the
body.
Membrane Stabilization.
The initial treatment
of hyperkalemia should be the infusion of calcium. Calcium antagonizes the
effects of hyperkalemia at the cellular level through 3 major mechanisms.
First, in the setting of hyperkalemia, the resting membrane potential is
shifted to a less negative value, that is, from [minus sign]90 mV to
[minus sign]80 mV, which in turn moves the resting membrane potential
closer to the normal threshold potential of [minus sign]75 mV, resulting
in increased myocyte excitability. When calcium is given, the threshold
potential shifts to a less negative value (that is, from [minus sign]75 mV
to [minus sign]65 mV), so that the initial difference between the resting
and threshold potentials of 15 mV can be restored.38 For
example, if a myocyte has a normal resting membrane potential of [minus
sign]90 mV and a normal threshold potential of [minus sign]75 mV, then 15
mV of depolarization is required before reaching the threshold potential.
In the setting of hyperkalemia, the resting membrane potential may change
to a new level (that is, [minus sign]80 mV), so that now only 5 mV of
depolarization must occur before reaching the threshold potential of
[minus sign]75 mV. When calcium is given, the threshold potential becomes
less negative (that is, it changes from [minus sign]75 mV to [minus
sign]65 mV). Thus the difference between the hyperkalemia-induced resting
membrane potential of [minus sign]80 mV and the calcium-induced threshold
potential of [minus sign]65 is now back to 15 mV, and myocyte excitability
can return to normal.
Second, it has been shown in animal studies that increasing levels of
calcium shift the curve relating Vmax to the resting membrane
potential at the onset of action potential upward and to the right (Fig.
5).9 Therefore, at any given level of resting membrane
potential, up to approximately [minus sign]75 mV, the Vmax is
increased when high calcium concentrations are present.39 This
serves to return myocyte excitability back to normal in the setting of
hyperkalemia, where myocyte depolarization is decreased secondary to
decreased rates of Vmax.
Finally, in cells with calcium-dependent action potentials, such as SA
and atrioventricular nodal cells, and in cells in which the sodium current
is depressed, an increase in extracellular calcium concentration will
increase the magnitude of the calcium inward current and the
Vmax by increasing the electrochemical gradient across the
myocyte. This would be expected to speed impulse propagation in such
tissues, reversing the myocyte depression seen with severe
hyperkalemia.40
The effects of intravenous calcium occur within 1 to 3 minutes but last
for only 30 to 60 minutes. Therefore, further, more definitive treatment
is needed to lower serum potassium levels. Calcium gluconate is the
preferred preparation of intravenous calcium. The dose should be 10 mL of
a 10% calcium gluconate solution infused over 2 to 3 minutes. Calcium
chloride may also be used but provides about 3 times the amount of calcium
per 10-mL dose, so the dose needs to be attenuated accordingly to avoid
potential calcium toxicity.36 Because hypercalcemia can
potentiate digitalis toxicity, calcium should be used in patients taking
digitalis only if there is loss of P waves or a widened QRS
complex.8 In this situation, calcium gluconate should be
diluted in 100 mL of D5W (dextrose [5%] in water) and infused
over 30 minutes.
Promotion of Potassium Influx into Cells. After the
administration of calcium, the next goal of treatment is to shift
potassium intracellularly. This is most frequently done by giving insulin.
Insulin stimulates the Na-K ATPase pump, which moves potassium
intracellularly in exchange for sodium in a 2:3 ratio; this effect is
independent of insulin's effect on glucose.41 Ten units of
intravenous insulin is typically given, followed by close monitoring of
serum blood sugar. Fifty mL of 50% dextrose is frequently co-administered
with insulin in normoglycemic patients to prevent hypoglycemia. If a
patient is already hyperglycemic, supplemental glucose is not needed. The
effect of the insulin is seen within 10 to 20 minutes of administration
and can be expected to decrease potassium levels by 0.6 to 1.0
mEq/L.36,42,43
Growing evidence suggests that there may be a role for albuterol in the
treatment of patients with severe hyperkalemia. Catecholamines activate
Na-K ATPase pumps through [beta]2 receptor stimulation in a
manner that is additive to the effect of insulin.36,44 In a
study by Montoliu and coworkers,41 0.5 mg of intravenous
albuterol was given to patients with hyperkalemia, leading to a 1-mEq/L
decrease in serum potassium levels with minimal adverse
effects.41 Because there are no approved intravenous forms of
[beta] agonists available in the United States, studies have been
performed to determine whether nebulized [beta] agonists would have a
similar effect on serum potassium levels. One such study found that
albuterol, when given in very high doses (10 --20 mg vs the normal 0.5
mg), decreased potassium levels by 0.62 to 0.98 mEq/L.45 The
onset of action for inhaled albuterol was immediate and lasted for 1 to 2
hours. Although in these studies the effects varied among individuals,
[beta]2 agonist administration was found to be safe and was
associated with a significant decrease in serum potassium levels.
Therefore, [beta]2 agonist therapy should be considered as an
adjunctive treatment for patients with severe hyperkalemia.
Sodium bicarbonate infusion can shift potassium from the extracellular
to intracellular space by increasing blood pH. However, routine
bicarbonate therapy for the treatment of hyperkalemia is
controversial.36,46 --49 In a study by Blumberg and
associates,50 12 dialysis patients with potassium levels of
5.25 to 8.15 mEq/L received 390 mmol of intravenous sodium bicarbonate
over a 6-hour period. No change in potassium levels was seen until 4 hours
after drug administration, when a decrease of 0.7 mEq/L was noted; at 6
hours, however, the decrease in potassi-um level was only 0.35
mEq/L.50 Due to the lack of a quick or sustained decrement in
potassium levels, physicians should reserve the use of intravenous sodium
bicarbonate for situations wherein severe acidemia is present or there is
another indication for its administration (such as phenobarbital or
tricyclic antidepressant overdose).
Potassium Removal from the Body.
The final
task in treating patients with severe hyperkalemia is to remove potassium
from the patient's body. The quickest, most efficient way to do this is
through the use of hemodialysis.42 In 1970, Morgan and
colleagues51 reported the removal of 48 mEq/L of potassium
using a Kiil dialyzer over a 10-hour period; others confirmed these
findings.52,53 Because of the time, expense, and invasive
nature of hemodialysis therapy, it is rarely used as a 1st-line treatment
for hyperkalemia unless a patient is already on dialysis and has
life-threatening hyperkalemia. For most patients, treatment with an
exchange resin such as sodium polystyrene sulfonate is more appropriate.
Ion exchange resins can be administered orally or rectally and work by
exchanging gut cations, most importantly potassium, for sodium ions that
are released from the resin. Most studies have found exchange resins to
decrease serum potassium levels by about 1 mEq/L over a 24-hour
period.54 It should be emphasized that the extended time
required for exchange resins to work exclude their use in the emergent
treatment of hyperkalemia. Exchange resins can cause significant
constipation and are typically given in combination with a laxative such
as sorbitol. Not only does a laxative prevent constipation, but it also
promotes the elimination of potassium from the gut once it binds to the
resin. Although generally safe, the combination of a resin and sorbitol
has been reported to cause intestinal necrosis, and as such should be used
cautiously and only when necessary.36,55,56