hyperkalemia in esrd

1134 Volume 6’ Number 4 - 1995

Hyperkalemia in End-Stage Renal Disease: Mechanismsand Management1’2Michael Allon3

M. Allon. Nephrology Research and Training Center,University of Alabama at Birmingham and VA MedicalCenter, Birmingham, Alabama

(J. Am. Soc. Nephrol. 1995; 6:1134-1 142)

ABSTRACTClinical investigations in the past few years hoveenhanced the understanding of the mechanisms ofhyperkalemia in patients with ESRD. The results ofthese studies have led to modifications In the acutetreatment and prevention of hyperkalemia in thispatient population. They have confirmed the efficacyof intravenous insulin, while raising doubts about theutility of intravenous bicarbonate, for the acute treat-ment of hyperkalemia. Moreover, the fradrenergicagonist albuterol has been shown to be a usefuladjunct to insulin for acutely lowering plasma potas-sium. Finally, there has been enhanced recognitionof nondietary factors that can predispose to hyperkal-emla In patients with ESRD, including prolonged fast-ing and the use of noriselective p-adrenergic block-ers. These new insights may improve the clinicalmanagement of hyperkalemla in patients with renalfailure.

Key Words: Potassium, insulin, albuterol, epinephrine, bicar-bonate

R ecent review articles have discussed in depth thepathogenesis and management of potassium

disorders in patients with ESRD (1,2). The goal of this

editorial review Is to focus primarily on recent clinicalinvestigations that have enhanced our understandingof potassium homeostasis in hemodialysls patientsand to consider their Implications for the treatmentand prevention of hyperkalemla in dialysis patients.

ACUTE TREATMENT OF HYPERKALEMIA IN ESRDPATIENTS

The definitive treatment of severe hyperkalemia inESRD patients is the removal of excess potassium byemergent hemodialysis. Because the initiation of dial-

ysIs often Involves a 1- to 2-h delay, more Immediate

1 Received February 7, 1995. Accepted April 26, 1995.

2The manuscript was presented in part at the 27th annual meeting of the

American Society ot Nephroiogy in Orlando (October 26-29, 1994).

3Corresponderrce to Dr. M. Ailon, Division of Nephroiogy, 668 LHR, 701 South 19thStreet, Birmingham, AL 35294.

1046.6673/0604-1 134$03.00/0Journal of the American Society of NephrologyCopyright 0 1995 by the American Society of Nephrology

temporizing measures are used. Mter the myocardiumis stabilized with intravenous calcium, therapy isdirected at decreasing plasma potassium acutely, by

shifting potassium from the extracellular to the intra-cellular fluid compartment. Until recently, the admin-istration of intravenous insulin and bicarbonate wasrecommended uniformly to promote this shift. Only inthe past few years have these textbook recommenda-tions been examined critically In carefully designed,

prospective studies. These investigations have con-firmed the efficacy of insulin, while casting doubts

about the utility of bicarbonate therapy. Moreover,these studies have established 13-adrenergic agonistsas useful adjuncts to insulin for the acute treatment ofhyperkalemla. The following section will examinesome of the studies that have led to a change in ourthinking.

To understand the rational treatment and preven-tion of hyperkalemia in ESRD patients. It is necessaryto understand the physiologic mechanisms that reg-ulate potassium homeostasls. Extracellular potas-sium concentration Is determined by two distinct

regulatory systems (3) (Figure 1). The first systemmaintains external potassium balance by regulatingpotassium excretion according to dietary intake. On atypical American diet containing approximately 100

mmol of potassium per day, 90 to 95 mmol Is excretedby the kidneys and 5 to 10 mmol is excreted by the

gut. The excretion of an acute potassium load is arelatively slow process, requiring several hours for

completion (4). In ESRD patients, there is an adaptiveIncrease in potassium excretion by the gut (5-9).Nevertheless, this adaptation is insufficient to com-pensate for the loss of renal excretory capacity.

The second homeostatic system maintains internalpotassium balance by regulating potassium shiftsbetween the intracellular and extracellular fluid com-partments. Internal balance can be achieved muchfaster than external balance: an acute potassium loadIs shifted into the cells within minutes (2). Total bodypotassium is about 3.300 mmol. Only 2% (about 65mmol) Is in the extracellular fluid compartment,whereas 98% Is in the intracellular fluid compart-ment, primarily in skeletal muscle and to a lesserextent In liver. This uneven distribution means thatrelatively small shifts of potassium from the intracel-lular to the extraceilular fluid compartment can pro-duce severe hyperkalemla. Conversely, relativelysmall potassium shifts from the extracellular to theintracellular fluid compartment can produce amarked decrease in plasma potassium. Extrarenalpotassium disposal (shifts from the extracellular to

the intracellular fluid compartment) assumes a criti-

-1 5-1#{149}

Allon

Journal of the American Society of Nephrology 1135

STOOL5-10��E�/day

Figure 1. Internal and external potassium balance In hu-

mans. Reproduced with permission from Ref. 3. Gl, gastroin-testinal; RBC, red blood cells.

cal role in the defense against hyperkalemia In ESRDpatients.

The two major physiologic factors that stimulateextrarenal potassium disposal are insulin and epi-nephrine. Flatman and Clausen (10) observed thestimulation of potassium Influx in isolated skeletalmuscle fibers incubated with insulin or epinephrine.Simultaneous exposure to both hormones producedan additive effect on potassium Influx, suggesting

different cellular mechanisms. These observationswould suggest that both insulin and epinephrine maybe useful for the treatment of hyperkalemia In dialysispatients. Whereas Insulin is routinely used for thispurpose in clinical practice. epinephrine is not.

The potassium-lowering effect of insulin in normalindividuals is well established and dose dependent; Itis increased progressively, going from moderate phys-iologic doses to high physiologic doses to phannaco-

logic doses (11). The effect of insulin on potassiumuptake Is distinct from that of eplnephrine. becausep-adrenergic blockade does not attenuate the de-crease in plasma potassium during insulin adminis-tration (12,13). Several studies have documented theefficacy of intravenous insulin in the acute treatmentof hyperkalemia in ESRD patients (14-17) (Figure 2).The magnitude of decrease in plasma potassium afterinsulin administration In hemodialysis patients iscomparable to that obtained in normal controls (14).Interestingly, the effect of insulin on cellular potas-sium uptake can be dissociated from its effect onglucose uptake. This dissociation was demonstratedIn an elegant clinical study that measured plasmapotassium In the brachial artery and vein of normalcontrols before and during a continuous infusion ofinsulin into the brachial artery (18). Insulin increasedthe arteriovenous gradIent for both potassium andglucose. However, the concomitant Infusion of

i� Plasma Potassium,mmol/L

-1 �O-

O 10 20 30 40 50 60Duration of Treatment, mm

B,carbonate

Ep�oephr.ne

insuhn/glucose

D,aiys�s

Figure 2. Changes in plasma potassium during the Intrave-nous infusion of 8.4% bicarbonate, epinephrine, or Insulin Inglucose and during hemodialysis. Values represent means± SE. Reproduced with permissIon from Ref. 16.

ouabaln with insulin prevented the stimulation ofpotassium uptake, without affecting glucose uptake.Thus, the effect of Insulin on potassium uptake ap-pears to be mediated by enhanced activity of the Na-Kpump, whereas the stimulation of glucose uptake isachieved by a membrane-bound glucose transporter(19). These two effects of insulin are often dissociatedIn hemodialysis patients. As a result, ESRD patientsare relatively resistant to the glucose-lowering effect ofinsulin (20-22), yet responsive to Its potassium-low-ering action (14).

When treating hyperkalemia, dextrose is routinelyadministered in conjunction with Insulin, to avoid theanticipated hypoglycemia. Because exogenous glu-cose stimulates endogenous insulin secretion, onemight be tempted to administer dextrose alone. Suchan approach is ill advised in the emergent situation. Ifendogenous insulin Is not secreted (as might occur ininsulin-dependent diabetics or in patients with inad-equate insulin reserve), the resulting hyperglycemia Islikely to aggravate the hyperkalemla (23-26). An acuteworsening of hyperkalemia Is not unique to hypergly-cemia; any maneuver that increases plasma osmolal-Ity promotes potassium shifts from the intracellular tothe extracellular fluid compartments. Thus, for exam-ple, acute Infusions of mannitol, hypertonic saline,and hypertonic arginine in dialysis patients also raiseplasma potassium acutely (27-29). The hyperkalemiceffect of hyperosmolality Is independent of changes inplasma insulin, cartecholamines, or acid-base status(28).

Epinephrlne at physiologic doses produces hypoka-lemla acutely In normal controls (30-34). The stimu-

30 60 90 120

Time, minutes TIme, minutes

Hyperkalemia in ESRD

1136 Volume 6- Number 4 - 1995

latlon of extrarenal potassium disposal by epineph-

rine is mediated by J3-2 adrenergic receptors. It isprevented by /3-2 selective blockers and by nonselec-tive /3 blockers, but not by /3-1 selective blockers(30.32,35). The ability of intravenous /3-2 agonists toinduce hypokalemla in normal IndivIduals (36,37)suggests their potential utility for acute therapy ofhyperkalemla in hemodlalysls patients. Montoliu eta!.(38) first reported that intravenous albuterol (0.5 mg),a /3-2 agonIst, acutely lowers plasma potassium in

hemodialysls patients. The effect was maximal (1.1mmol/L reduction) at 30 mm and was sustained for atleast 3 h. In several patients, hyperkalemic electrocar-diographic changes resolved after the administrationof albuterol alone, in the absence of Intravenous cal-cium or any other measures to lower plasma potas-sium. Subsequent studies have confirmed the efficacyof Intravenous albuterol In the acute treatment ofhyperkalemla In patients with renal failure (17,39,40).

Neither albuterol nor any other /3-2 selective agonistIs available for Intravenous use In the United States.However, selective (3-2 agonists are available in nebu-

lized form for the treatment of asthma. Moreover,nebullzed albuterol (and other /3-agonists) produces

hypokalemla in normal controls (4 1-44). Indicatingsignificant systemic absorption of the inhaled drug. Todetermine whether nebulized albuterol was useful forthe acute treatment of hyperkalemla in ESRD pa-tients, we studied ten hemodialysis patients In adouble-blinded, crossover trial (45). Each subject re-ceived 10 or 20 mg of nebulized albuterol or nebulizedplacebo (saline) inhaled over 10 mm. Plasma potas-

slum decreased by about 0.6 mmol/L after 10 mg ofalbuterol and by about 1.0 mmol/L after the 20-mgdose (Figure 3). A significant decrease In plasma po-tassium was first noted after 30 mm. Albuterol wasequally efficacious in diabetic and nondlabetic pa-tients. In contrast, nebulized placebo did not lowerplasma potassium. Subsequent studies have con-firmed the utility of nebulized albuterol (10 to 20 mg)In lowering plasma potassium acutely (15.40,46).

The doses of albuterol required to achieve a signifi-

cant reduction in plasma potassium are fourfold to

eightfold higher than those used to treat asthma.raising concerns about potential adverse cardiovascu-lar effects. The mean increase in heart rate in ourdouble-blinded study was 6 beats/min with 10 mg ofalbuterol and 8 beats/min wIth 20mg of albuterol (45)(Figure 3). There were no significant changes in bloodpressure, and no patient developed chest pain orarrhythmias. Several subsequent investigations havesimilarly reported stable blood pressures and clini-cally insignificant Increases in heart rate after thenonblinded administration of albuterol at similar neb-ulized doses to ESRD patients (15,40,46). An increasein plasma glucose after intravenous and nebulizedalbuterol has been observed in a number of studies,but this has not been clinically significant (‘-2 to 3mmol/L).

A direct comparison between intravenous (0.5 mg)and nebullzed (10 mg) albuterol in ESRD patientsrevealed a similar magnitude of decrease in plasmapotassium (40). The onset of the potassium-loweringeffect was somewhat slower with the nebulized drugthan with the intravenous form. On the other hand,nebulized aibuterol produced less tachycardia than

the intravenous form. Thus, nebulized albuterol maybe safer than the intravenous form in patients withknown or suspected coronary artery disease.

It Is Important to note that a subset of ESRD pa-tients are refractory to the potassium-lowering effectof albuterol (�K � 0.4 mmol/L). In various studies,these have ranged from 20 to 40% of dialysis patients

studied (15,40.45). Similarly. even pharmacologicdoses of epinephrine do not lower plasma potassiumin some dialysis patients (47.48). The reason for thisresistance has not been elucidated. Because we can-not prospectively Identify which patients are resistantto albuterol, It Is critical to treat severe hyperkalemlawith intravenous insulin, as well as albuterol.

The potassium-lowering effect of albuterol in ESRD

patients Is additive to that of insulin. We found thatIntravenous insulin (10 U bolus) and albuterol (20 mgnebulized over 10 min) each lowered plasma potas-sium by approxImately 0.6 mmol/L, whereas in com-bination, they lowered plasma potassium by about 1.2

-J

I

IFigure 3. The effect of nebulized albuterol in doses of 10 mg (solid bar) or 20 mg (stippled bar) or of nebulized placebo (openbar) on the plasma potassium concentration (left panel) and heart rate (right panel) of hemodlalysis patients. Values representmeans ± SE. P < 0.05 versus nebulized placebo. �p < 0.05 versus 10 mg of albuterol. Adapted from Ref. 45.

“Pure” Alpha0.8

0.6

0.4

0.2

0.0 F0.01 0.02 0.04

0EEE

to0a.

a,cmCto

C0

-J

0EE

EtnU,‘a

00.

C

a,cmCto

r0

Allon


mmol/L (15). A similar additive effect has also beenreported with intravenous albuterol and Insulin (17).This additive effect is consistent with the in vitro

observations in isolated skeletal muscle fibers (10).Albuterol may also attenuate insulin-induced hypo-

glycemia in hemodialysis patients. Dextrose is admin-istered routinely with intravenous Insulin to prevent

the predictable hypoglycemia. Nevertheless, hypogly-cemia (frequently asymptomatic) may be a commonevent after such therapy. For example, we docu-mented hypoglycemia (plasma glucose < 3 mmol / L) in9 of 12 hemodialysis patients 1 h after they received10 U of regular insulin and 25 g of dextrose as an

intravenous bolus (15). The high incidence of hypogly-cemia may reflect, in part, a defect in the activation ofcounterregulatory sympathetic discharge m ESRD pa-tients (49). When nebulized albuterol was given con-currently with the same regimens of insulin and dex-trose, only 2 of 10 patients developed hypoglycemia.This salutory action of albuterol is likely mediated bythe $3-adrenergic stimulation of gluconeogenesis (50).

In contrast to /3-adrenergic stimulation, which en-hances potassium influx into the cells, a-adrenergicstimulation enhances potassium efflux out of the cells(51.52). Hence, the effect of epinephrine, a mixed a-

and /3-blocker, on plasma potassium reflects the rela-tive effects of a- and /3-adrenergic stimulation onplasma potassium. Hemodialysls patients are resis-

tant to the potassium-lowering effect of epinephrlne.For example, Gifford et al. (34) found that a low

physiologic dose of epinephrine (0.015 �g/ kg per min)lowered plasma potassium by about 0.4 mmol/L Innormal controls but had no effect in hemodialysispatients. Similarly, Blumberg et at. (16) observed nosignificant change In plasma potassium during theinfusion of epinephrine at high physiologic doses (0.05pg/kg per min) in hemodIalysis patients (Figure 2).These findings are in contrast to the potassium-low-ering effects of /3-2 agonists in ESRD patients, sug-gesting a possible imbalance between a- and /3-adren-ergic effects on potassium In dialysis patients.

To evaluate this possibility, we gave ESRD patients

and normal controls a graded Infusion of epinephrmneat three physiologic doses (0.01 to 0.04 ,.tg/kg per mm)(33). In normal controls, increasing doses of epmneph-rifle produced progressively greater decreases in

plasma potassium. In contrast, in hemodialysis pa-

tients, epmephrine at the lowest dose actually in-creased plasma potassium, suggesting a predominanta-adrenergic effect. Even the highest dose of epineph-nine did not significantly decrease plasma potassiumin ESRD patients (Figure 4). To indirectly assess the

a-adrenergic component of epinephrine, the identicalinfusion protocol was repeated during concurrent /3blockade with propranolol. a-Adrenergic stimulationraised plasma potassium in a dose-dependent man-ner in both experimental groups. The magnitude ofthis increase was significantly greater. however, indialysis patients (0.7 mmol/L) than in normal controls(0.3 mmol/L). Finally, the concomitant infusion of the

Alpha + Beta U Control�? 0.41

U Dialysis

E� ___0.2 � #{149}Transplant

J‘� II!acm

-0.80.01 0.02 0.04

0.01 0.02 0.04

Epinephrine dose (�ig/kgJmin)

Figure 4. Changes in serum potassium concentration(means ± SE) from baseline during the infusion of Intrave-nous epinephrine at three doses in normal controls, hemo-dialysIs patients, and renal transplant patients (serum creat-mine <2 mg/mi). The upper panel shows the effect ofepinephrlne alone (a- plus f3-adrenerglc stimulation), the

middle panel shows the effect of epmnephrlne during $3blockade (‘pure’ a effect), and the lower panel shows thecalculated pure /3 effect. The changes in potassium in theupper and middle panels were signifIcantly (P < 0.05)different In the dialysis patients as compared with those inthe controls and transplant patients. Reproduced with per-mission from Ref. 2.


1138 Volume 6� Number 4’ 1995

a-blocker phentolamine overcame the resistance to

the potassium-lowering effect of epinephrine in thedialysis patients (33). In summary, these studies sug-gest that the resistance to the potassium-loweringeffect of epmnephrmne In hemodlalysis patients Is due to

an enhanced a-adrenerglc response. Interestingly,this acquired resistance Is reversed after successfulrenal transplantation (2) (FIgure 4).

Bicarbonate therapy has long been assumed to beeffective for the acute treatment of hyperkalemia Indialysis patients. It is presumed that the metabolicacidosls of renal failure promotes hyperkalemia by

stimulating net potassium shifts from the intracellu-lar to the extracellular fluid compartment and thatalkallnization should reverse this action. However, theeffect of acute acid infusion on plasma potassium isInconsistent. In nephrectomlzed animals, the acute

(1- to 3-h) administration of mineral acids raises

plasma potassium, whereas organic acids have noeffect (53-56). The effect of acute acid administrationon plasma potassium in dialysis patients has not beenreported. Acute (1- to 3-h) bicarbonate administrationto nephrectomlzed animals has lowered plasma potas-sium in some studies (56,57), but not in others

(54,55). SodIum bicarbonate infusion in hyperkalemicdialysis patients lowers plasma potassium after 6 to24 h (58). However, to be useful for acute therapy, thechanges should occur within about 1 h. Blumberg et

a!. (16) found that neither hypertonic nor isotonic

intravenous bicarbonate lowered plasma potassiumin hemodlalysis patients during 60 mm of observa-tion. This was in contrast to the abifity of both insulinand hemodlalysis to lower plasma potassium (Figure

2). Even 3 h after intravenous bicarbonate (hypertonicor Isotonic) administration, there was no significantdecrease in plasma potassium in ESRD patients, de-

spite substantial Increases in plasma bicarbonate

(59). In fact, serial plasma potassium measurementsfor 6 h after bicarbonate administration found theearliest decrease at 4 h (60). In summary, although

bicarbonate administration lowers plasma potassiumchronically In ESRD patients. It does not do so in atimeframe useful for the emergent treatment of hy-perkalemia. Of course, bicarbonate therapy Is stillindicated for the acute management of severe meta-bolic acidosis.

A recent abstract suggested that bicarbonate ad-ministration may potentiate the potassium-loweringaction of Insulin in hemodialysis patients (61).

Whereas bicarbonate administration by Itself had no

significant effect on plasma potassium after 60 min Ineight hemodialysis patients, It enhanced the potassi-um-lowering effect of insulin. Plasma potassium wasdecreased by 0.6 mmol/L after insulin alone and by1.0 mmol/L after insulin plus bicarbonate. The cellu-lar mechanism of this synergIsm Is unexplained butmay Involve pH-dependent changes in the insulinreceptor. If this study Is confirmed, It may provide arationale for the administration of bicarbonate for theacute treatment of hyperkalemla in ESRD patients.

Whether a similar synergism exists between bicarbon-ate and (3-2 adrenerglc agonists is unknown.

The removal of potassium by hemodialysis is largelydetermined by the potassium concentration gradientbetween the plasma and the dlalysate (62,63). Mea-

sures that acutely shift potassium from the extracel-

lular to the intracellular fluid compartment will re-duce this gradient and thereby attenuate potassiumremoval by the subsequent dialysis treatment. Suchan effect of insulin is suggested by studies demon-strating less dialytic potassium removal with glucose-containing dialysate as compared with glucose-freedialysis (64). Thus, one might anticipate that aggres-sive therapy with insulin and albuterol for hyperkale-mia may attenuate the removal of potassium in thesubsequent dialysis treatment. Such a phenomenonmight potentially lead to a rebound hyperkalemia

several hours after the dialysis session.

PREVENTION OF HYPERKALEMIA IN ESRD PATIENTS

Given the central role of the kidneys in maintainingpotassium homeostasis. dietary potassium restriction

clearly plays an Important role in the prevention ofhyperkalemia. As nephrologists, we tend to assumedietary excess whenever we encounter hyperkalemiain hemodialysis patients. We then ask the dietitians to

provide the patients with appropriate dietary guide-lines. It Is important, however, to recognize somenondietary factors that may predispose to hyperkale-mia in hemodialysls patients.

Thus, for example, there is convincing evidence forthe impairment of extrarenal potassium disposal inESRD patients. In a recent study, we gave fastedhemodialysis and control subjects a small oral potas-slum load (0.25 mmol/kg, or about 20 mmol for an80-kg patient). The mean increase in plasma potas-sium was substantially greater In hemodialysis pa-

tients than in controls (0.8 versus 0.4 mmol/L) and

could not all be accounted for by renal excretion (65).Similar observations were reported by Fernandez et at.

(66). These clinical observations in ESRD patients arealso in agreement with the impairment of ouabamn-inhibitable rubidium uptake by skeletal muscle fromuremic rats (67,68).

Patients do not usually eat potassium by itself. Theyeat foods that contain potassium, as well as carbohy-

drate. The endogenous insulin that is released in

response to the carbohydrate promotes extrarenalpotassium disposal. We found that the ingestion of 50g of glucose along with an oral potassium load (0.25mmol/kg) attenuated the mean rise In plasma potas-

sium from 0.8 to 0.3 mmol/L (65). If one assumes thatthe extracellular fluid compartment is approximately20% of total body water and that all the ingestedpotassium (0.25 mmol/kg) remains in the extracellu-lar fluid compartment, then one would predict a 1.25mmol/L increase in plasma potassium. The actualmeasured increases suggest that endogenous insulinapproximately doubles extrarenal potassium dis-

+

4.80E 4.6,g. �

E� 4.2U,U,U,

0a. 3.8U,

E 3.60).! 3.40. 0

�30,

.E 20

0,C

I i:

-I #{149} I

6 12 18

V

0 18

Allon


posal. from 36 to 76%. Thus, dietary carbohydrateconstitutes a major defense against life-threatening

hyperkalemia after an oral potassium load in dialysis

patients.The ability of exogenous insulin to lower plasma

potassium acutely has been discussed earlier. What Is

less well appreciated is that even the low circulating

insulin concentrations present during prolonged fast-ing constitute an important defense against hyperkal-emia. Thus, the Inhibition of endogenous insulin se-cretion in normal, fasting controls results in a

significant increase in plasma potassium (69). Fastinginsulin levels may be particularly important in the

defense against hyperkalemia in dialysis patients.This conclusion is suggested by the finding that acutesomatostatin administration to fasting rats produceda greater increase in plasma potassium in uremic ratsthan in control animals (68).

The decrease in plasma insulin that occurs in thetransition from a postprandial state to a fasting stateis not sufficient to raise plasma potassium In Individ-

uals with normal renal function (70). Presumably, theexcess potassium that is shifted from the intracellularto the extracellular fluid compartment is excreted bythe kidneys. In the absence of renal excretory func-tion, however, prolonged fasting may predispose to

hyperkalemia. Thus, Gifford et a!. (34) fasted nondla-betic hemodlalysis patients and normal controls for26 h. Plasma potassium increased by about 0.5mmol / L in ESRD patients but did not change in thecontrols. This observation suggests that prolongedfasting may be an Important cause of hyperkalemiawhen hemodialysis patients are fasted for surgery or a

diagnostic procedure.

To determine whether fasting hyperkalemia couldbe prevented by exogenous insulin, we fasted 10nondiabetic hemodialysis patients for 18 h and mea-

sured serial plasma potassium concentrations (70).

On a second occasion, the same patients received acontinuous infusion of 10% dextrose with 20 U ofregular insulin per liter, infused at 50 mL/min (or 1 Uof insulin/h). This insulin infusion regimen doubledfasting plasma insulin from 10 to 20 mU/L and com-

pletely prevented the 0.6 mmol/L increase in potas-

sium observed during fasting (Figure 5). Even low

doses of exogenous insulin can potentially producehypoglycemia. To evaluate whether exogenous glu-cose can prevent fasting hyperkalemla. by stimulatingendogenous Insulin secretion, we restudied some pa-tients with an infusion of 10% dextrose alone. Thisattenuated the rise in plasma potassium during pro-longed fasting but was not as effective as dextrose withinsulin (70).

Hyperkalemla is much less common In patients on

continuous ambulatory peritoneal dialysis than Inhemodialysis patients (71). This Is due, in part, to thecontinuous removal of potassium with continuous

ambulatory peritoneal dialysis. However, another con-tributory factor Is the continuous absorption of glu-cose from the dialysate, which results in increased

I #{149}

6 12Hours of Fasting

Figure 5. Serial plasma potassium and Insulin concentrationsduring an 18-h fast in hemodlalysis patients In the absence(solid squares) or presence (open squares) of a continuousphysiologic infusion of insulin (1 U/h) with dextrose (5 glh).Values represent means ± SE for 10 patIents. ‘p < 0.05 and“P < 0.01 versus the value at 6 h of fasting; �P < 0.01 versusthe corresponding value In the control experiment. Repro-duced with permission from Ref. 70.

fasting plasma Insulin concentrations, thereby pro-moting extrarenal potassium disposal (72,73).

Because /3-2 agonists decrease plasma potassium,one might expect /3-blockers to increase plasma potas-

sium. This is not a significant problem in individualswith normal renal function, because potassium trans-ferred from the intracellular to the extracellular fluidcompartment can be excreted by the kidneys. In theabsence of renal excretory function, /3 blockade can

aggravate hyperkalemia in ESRD patients. Thus, forexample, a 10-day course of propranolol, a nonselec-tive /3-blocker, increased predlalysis potassium by 0.7

mmol/L In a group of hemodialysis patients. In con-trast, the /3-1 selective blocker atenolol did not in-crease plasma potassium in the same group of pa-

tients (74). Similarly, exercise-induced hyperkalemlain ESRD patients is augmented by the nonselective

0.6

*

*


1140 Volume 6’ Number 4- 1995

/3-adrenergic blocker propranolol but is not affected bythe /3-1 selective blocker metoprolol (75) (Figure 6).Thus, one should preferentially prescribe /3-1 selectIveblockers in dialysis patients.

A recent abstract described the potential utility ofchronic oral 13-adrenerglc administration (albuterol, 2mg twice daily) In the prevention of hyperkalemla in

hemodialysis patients (76). The predialysis plasma

potassium was decreased by 0.6 mmol/L after 2 wk of

oral albuterol. However, 3 of the 12 patients withdrewfrom the study because of unacceptable palpitations

and tremor. Further studies are indicated to evaluatethis prophylactic measure In hemodialysis patientswith persistent hyperkalemia not responsive to di-etary potassium restriction.

Mlneralocorticolds play an important role In main-

taining potassium homeostasis In normal animals

and humans fed a high-potassium diet by enhancingrenal potassium excretion (77,78). In addition, there isconflicting experimental evidence on the effect of mm-eralocorticolds on extrarenal potassium disposal (2).Sugarman and Brown (79) studIed the effect of

i. chronic mineralocortlcold administration on the dis-

0.5

0.4

0.3

Ut

0.2

0.1

0.0

-0.1

-0.2

-0.3

Exercise Recovery

0 10 20 30 40 50 60 70

Time, minutes

Figure 6. Time course of change in plasma potassium con-centration in dialysis subjects during exercise alone (opencircles), exercise plus propranolol (closed circles), and ex-ercise plus metoprolol (closed triangles). All values representmean ± SE. ‘P < 0.05 versus control study. P < 0.05 versusmetoprolol study. Reproduced with permission from Ref. 75.

TABLE 1. Changing concepts in the management ofhyperkalemia in dialysis patients

Conventional Wisdom ‘New’ Wisdom

Bicarbonate decreases Bicarbonate is not effectiveplasma potassium for acute therapy ofacutely. hyperkalemla.

- Albuterol decreasesplasma potassiumacutely.

Insulin decreases plasma Insulin decreases plasmapotassium acutely. potassium acutely, and

albuterol has an additiveeffect.

Hyperkalemia is due to Hyperkalemia may be duedietary excess. to dietary excess, but

fasting and $3-blockersare also importantfactors.

TABLE 2. Treatment and prevention of hyperkalemiain dialysis patients: specific guidelines

Acute Treatment of Severe Hyperkalemia1. Calcium gluconate, 10% solutIon, 10- to 20-mi iv

bolus. May be repeated every 5 mm, ifelectrocardiographic appearance does not improve.

2. Regular insulin, 10 U + 50 ml of 50% dextrose, as ivbolus, followed by 10% dextrose 0 50 mL/mln untildialysis initiated.

3. Albuterol (5 mg/mi), 10 to 20 mg, nebulized over 10mm.

4. Acute hemodialysis against a low potassium bath.

Prevention of Hyperkalemia1. Dietary potassium restriction, 40 to 50 mmoi/day or

1.5 to 2.0 g/d.2. Avoidance of nonselective /3-blockers. If 13-blockers

are clinically indicated, use /3-1-selective blockers,e.g., atenolol or metoprolol.

3. During prolonged fasting in diabetic dialysis patients,infuse: 10% dextrose + 10 U of regular insulin/I 0 50mi/h.During prolonged fasting in nondiabetic dialysispatients, infuse: 10% dextrose 0 50 mL/h.

posal of an acute potassium load in hemodialysispatients. The prior exogenous administration of themineralocorticoid deoxycorticosterone acetate for 60 h

blunted the increase in plasma potassium after anacute potassium load. Conversely, the prior adminis-

tration of the aldosterone antagonist spironolactoneaugmented the hyperkalemic response to the potas-sium load. These intriguing observations suggest thatexogenous mineralocorticoids may be a useful modal-ity for the prevention of hyperkalemia in hemodialysispatients.

In summary, clinical investigations performed In thepast few years have enhanced our understanding of

Alion


the mechanisms of hyperkalemia in ESRD patients.As a result, the guidelines for the acute treatment and

prevention of hyperkalemia in hemodlalysis patientsmay need to be revised (Tables 1 and 2). The utility of

intravenous calcium and insulin for acute hyperkale-

mia associated with electrocardiographic abnormali-ties remains unchallenged. In contrast, the efficacy of

intravenous bicarbonate for the treatment of hy-perkalemia is not supported by prospective studies.

Albuterol, in either intravenous or nebulized form,appears to be a useful adjunct to the potassium-lowering action of intravenous insulin. Finally,whereas dietary potassium restriction remains anImportant measure for the prevention of hyperkale-mia, we have become more aware of nondietary factorscontributing to hyperkalemia, including prolongedfasting and nonselective /3-blockers.

ACKNOWLEDGMENTS

The author gratefully acknowledges the contributions of the following

individuals to his Investigations of potassIum homeostasis In ESRDpatients: Nancy Shanklin. Linda Dansby, Azlta Takeshian. Charles

Copkney, and Robert Dunlay.

REFERENCES

1. Salem MM, Rosa RM, Baffle DC: Extrarenal potassiumtolerance In chronic renal failure: Implications for thetreatment of acute hyperkalemia. Am J Kidney DIs 1991;18:42 1-440.

2. Allon M: Treatment and prevention of hyperkalemia inend-stage renal disease. Kidney mt 1993:43:1197-1209.

3. DeFronzo RA, Bia M. Extrarenal potassium homeosta-sis. In: Seldin DW, Gleblsch G. Eds. The Kidney. Physi-ology and Pathology. New York: Raven Press; 1985:1179-1206.

4. Kurtzman NA, Gonzalez J, DeFronzo RA, Giebisch G: Apatient with hyperkalemla and metabolic acidosis. Am JKidney Dis 1990;15:333-356.

5. Hayes CP, McLeod ME, Robinson RE: An extrarenalmechanism for the maintenance of potassium balance Insevere chronic renal failure. Trans Assoc Am Phys 1967;80:207-2 16.

6. Basil C, Hayslett JP. Binder HJ: Increased large Intesti-nal secretion of potassium in renal insufficiency. Kidneymt 1977;12:9-16.

7. Sandle GI, Galger E, Tapster S. Goodship THJ: Evidencefor large Intestinal control of potassium homeostasis Inuremic patients undergoing long-term dialysis. Clin Sd1987:73:247-252.

8. Panese S. Martin RS, Virginillo M, Litardo M, Arrizur-ieta E, Hayslett JP: Mechanism of enhanced transcellu-lar potassium secretion In man with chronic renal fail-ure. Kidney mt 1987;3 1:1377-1382.

9. Martin RS, Panese S. Virginillo M, et at.: Increasedsecretion of potassium in the rectum of humans withchronic renal failure. Am J Kidney Dis 1986:8:105-110.

10. Flatman JA, Clausen T: Combined effects of adrenalineand Insulin on active electrogenlc Na+-K+ transport Inrat soleus muscle. Nature 1979:281:580-581.

11. DeFronzo RA, Felig P. Ferrannini E, Wahren J: Effect ofgraded doses of Insulin on splanchnlc and peripheralpotassium metabolism In man. Am J Physlol 1980:238:E421-E427.

12. Massara F, Camanni F. Molinatti G: Effect of propranololon some adrenaline- and Insulin-induced metabolicchanges In man. Acta Diabetol Lat 1975:12:41-51.

13. Mlnaker KL, Rowe JW: Potassium homeostasis dunnhyperinsulinemia: effect of insulin level, b-blockade, anage. Am J Physiol 1982:242:E373-E377.

14. Alvestrand A, Wahren J, Smith D, DeFronzo RA: Insu-

lin-mediated potassium uptake is normal in uremic andhealthy subjects. Am J Physiol 1984;246:El74-E180.

15. Allon M, Copkney C: Albuterol and insulin for treatmentof hyperkalemia in hemodlalysIs patients. Kidney Int1990:38:869-872.

16. Blumberg A, Weidinann P, Shaw S. Gnadinger M: Effectof various therapeutic approaches on plasma potassiumand major regulating factors In terminal renal failure.Am J Med 1988:85:507-5 12.

17. Lens XM, Montollu J, Cases A, Campistol JM, Revert L:Treatment of hyperkalemia In renal failure: Salbutamol vInsulin. Nephrol Dial Transplant 1989:4:228-232.

18. Ferrannini E, Taddei S. Santoro D, et aL: Independentstimulation of glucose metabolism and Na+-K+ ex-change by insulin in the human forearm. Am J Physlol1988;255:E953-E958.

19. Hirshman MF, Goodyear U, Wardzala U, Horton ED,Horton ES: Identification of intracellular pool of glucosetransporters from basal and insulin-stimulated rat skel-etal muscle. J Biol Chem 1990:265:987-99 1.

20. SmIth D, DeFronzo RA: Insulin resistance in uremiamediated by postblnding defects. Kidney mt 1982:22:54-62.

21. Alvestrand A, Mujagic M, Wajngot A, Efendlc 5: GlucoseIntolerance In uremic patients: The relative contribu-tions of Inpaired beta-cell function and Insulin resis-tance. Clin Nephrol 1989:31:175-183.

22. Hager SR: Insulin resistance of uremia. Am J Kidney Dis1989:14:272-276.

23. Vlberti GC: Glucose-induced hyperkalemia: A hazard fordiabetics? Lancet 1978:1:690-691.

24. Nicolis GL, Kahn T, Sanchez A, Gabrllove JL: Glucose-Induced hyperkalemla In diabetic subjects. Arch InternMed 1981:141:49-53.

25. Goldfarb S. Cox M, Singer I, Goldberg M: Acute hy-perkalemia Induced by hyperglycemia: Hormonal mech-anisms. Ann Intern Med 1976:84:426-432.

26. Sunderlin FS, Anderson GH, Streeten DHP, BlumenthalSA: The renin-angiotensin-aldosterone system In dia-betic patients with hyperkalemia. Diabetes 1981:30:335-340.

27. Moreno M, Murphy C. Goldsmith C: Increase in serumpotassium resulting from the administration of hyper-tonic mannitol and other solutions. J Lab Clin Med1969:73:291-298.

28. Conte G, Dal Canton A, Iniperatore P. et at.: Acuteincrease in plasma osmolality as a cause of hyperkale-mia In patients with renal failure. Kidney Int 1990:38:301-307.

29. Hertz P. Richardson JA: Arginine-induced hyperkalemiain renal failure patients. Arch Intern Med 1972:130:778-780.

30. Brown MJ, Brown DC, Murphy MB: Hypokalemla frombeta2-receptor stimulation by circulating eplnephrlne. NEngl J Med 1983:309:1414-1419.

31. Brown RS: Extrarenal potassium homeostasis. KidneyInt 1986:30:116-127.

32. Struthers AD, Reid JL: Adrenaline causes hypokalemlaIn man by beta-2 adrenoceptor stimulation. Clin Endo-crlnol 1984:20:409-414.

33. Allon M, Shanklin N: Adrenergic modulation of extrare-nal potassium disposal In men with end-stage renaldisease. Kidney Int 1991:40:1103-1109.

34. Gifford JD, Rutsky EA, Kirk KA, McDanIel HG: Controlof serum potassium during fasting In end-stage renaldisease. Kidney Int 1989:35:90-94.

35. Struthers AD, Reid JL, Whitesmith R. Rodger JC: Theeffects of cardioselective and non-selective beta-adrenoceptor blockade on the hypokalemic and cardio-vascular responses to adrenomedullary hormones inman. Clin Sd 1983:65:143-147.

36. VIncent HH, Boomsma F. Man in ‘t Veld AJ, Derkx FHM,Wenting GJ, Schalekamp MADH: Effects of selective andnonselective beta-agonists on plasma potassium andnoreplnephrine. J Cardlovasc Pharmacol 1984:6:107-114.

37. Nevifie A, Palmer JBD, Gaddle J, May CS, Palmer KJ�V,Murchison LE: Metabolic effects of salbutamol: Compar-

Hyperkalemia in ESPD

1142 Volume 6 . Number 4 . 1995

Ison of aerosol and Intravenous administration. Bnit MedJ 1977:1:413-414.

38. Montoilu J, Lens XM, Revert L: Potassium-loweringeffect of albuterol for hyperkalemla in renal failure. ArchIntern Med 1987:147:713-717.

39. Murdoch IA: Dos Anjos R, Haycock GB: Treatment ofhyperkalemla with intravenous salbutamol. Arch DisChild 1991:66:527-528.

40. Liou HH, Chiang SS, Wu SC, et at.: Hypokalemic effectsof Intravenous Infusion or nebulizatlon of salbutamol Inpatients with chronic renal failure: Comparative study.Am J Kidney DIs 1994:23:266-271.

41. Lipworth BJ, Tregaskis BF, McDevitt DG: Beta-adrenoceptor responses to inhaled salbutamol in theelderly. Br J Clin Pharmacol 1989:28:725-729.

42. Lipworth BJ. McDevitt DG, Struthers AD: Prior treat-ment with diuretic augments the hypokalemic and elec-trocardiographic effects of Inhaled albuterol. Am J Med1989:86:653-657.

43. Haalboom JRE. Deenstra M. Struyvenberg A: Hypoka-lemla Induced by Inhalation of fenoterol. Lancet 1985:1:1125-1127.

44. Wong CS, Pavord ID, Williams J, Britton JR. Tatters-field AE: Bronchodilator, cardiovascular, and hypokale-mid effects of fenoterol. salbutamol, and tenbutaline Inasthma. Lancet 1990:336:1396-1399.

45. Allon M. Dunlay R, Copkney C: Nebullzed albuterol foracute hyperkalemla in patients on hemodlalysis. AnnIntern Med 1989:110:426-429.

46. Montollu J, Almirall J, Ponz E. Campistol JM, Revert L:Treatment of hyperkalemla in renal failure with salbuta-mol inhalation. J Intern Med 1990:228:35-37.

47. MartInez Vea M, Montollu J, Andreu L, et aL: Beta-adrenerglc modulation of extrarenal potassium disposalin terminal uremia. Proc EDTA 1982:19:756-760.

48. Yang WC, Huang TP. Ho LT, Chung HM, Chan YL, BaffleDC: Beta-adrenergic-medlated extrarenal potassiumdisposal in patients with end-stage renal disease: Effectsof propranolol. Miner Electrolyte Metab 1986:12:186-193.

49. Adrogue HJ: Glucose homeostasls and the kidney. Kid-ney Int 1992:42:1266-1282.

50. Rizza RA, Cryer PE, Haymond MW, Gerich JE: Adren-ergic mechanisms for the effects of epinephrIne on glu-cose production and clearance In man. J Chin Invest1980:65:682-689.

51. Williams ME, Rosa RM. Silva P. Brown RS. Epstein PH:Impairment of extrarenal potassium disposal by alpha-adrenerglc stimulation. N EngI J Med 1984:311:145-149.

52. Williams ME. Gervino EV, Rosa RM, et aL: Catechol-amine modulation of rapid potassium shifts during ex-ercise. N EngI J Med 1985:312:823-827.

53. Swan RC, Pitts RF: Neutralization of Infused acid bynephrectomized dogs. J Chin Invest 1955:34:205-212.

54. Tobln RB: Varying roles of extracellular electrolytes Inmetabolic acidosis and alkalosis. Am J Physlol 195:1958.685-692.

55. Poole-Wilson PA. Cameron IR: Intracellular pH and K+of cardiac and skeletal muscle in acldosis and alkalosls.Am J Physlol 1975:229:1305-13 10.

56. Keating RE. Weicbselbaum TE. Alanis M. Margraf HW,Elman R: The movement of potassium during experi-mental acidosis and alkalosis in the nephrectomlzeddog. Surg Gynecol Obstet 1953:96:323-330.

57. Swan RC, Axelrod DR. Seip M, Pltts RF: Distribution ofsodium bicarbonate infused into nephrectomlzed dogs. JClin Invest 1955:34:1795-1801.

58. Fraley DS. Adler 5: Correction of hyperkalemla by bicar-bonate despite constant blood pH. Kidney mt 1977:12:354-360.

59. Guttierez R, Schlesslnger F, OsterJR, Rietberg B, PerezGO: Effect of hypertonic versus Isotonic sodium bicar-bonate on plasma potassium concentration in patients

with end-stage renal disease. Miner Electrolyte Metab1991:17:297-302.

60. Blumberg A, Weidmann P. Ferrari P: Effect of prolongedbicarbonate administration on plasma potassium in ter-minal renal failure. Kidney Int 1992:41:369-374.

61. Kim HJ. Park CH. Ahn YH, Kang CM, Park HC: Syner-gistic effect of sodium bicarbonate and insulin In glucosein the therapy of hyperkalemia accompanied by meta-bolic acidosis [Abstract). J Am Soc Nephrol 1994:5:369.

62. Williams AJ, Barnes JN. Cunningham J, Goodwin FJ,Marsh FP: Effect of diahysate buffer on potassium re-moval during hemodlalysis. Proc Eur Dial Trans Assoc1984:21:209-214.

63. Sherman RA, Hwang ER. Bernholc AS, Eisinger RP:Variability in potassium removal by hemodlalysis. Am JNephrol 1986:6:284-288.

64. Ward RA, Wathen RL, Williams TE, HardIng GB: Hemo-dialysate composition and intradlalytlc metabolic, acid-base and potassium changes. Kidney Int 1987:32:129-135.

65. Allon M, Dansby L, Shanklin N: Glucose modulation ofthe disposal of an acute potassium load In patients withend-stage renal disease. Am J Med 1993:94:475-482.

66. Fernandez J, Oster JR. Perez GO: Impaired extrarenaldisposal of an acute oral potassium load In patients withendstage renal disease on chronic hemodlalysis. MinerElectrolyte Metab 1986:12:125-129.

67. Bonilla S. Goecke IA, Bozzo 5, Alvo M, Michea L,Marusic ET: Effect of chronic renal failure on Na,K-ATPase alpha-i and alpha-2 mRNA transcription in ratskeletal muscle. ,.J Chin Invest 1991:88:2137-2141.

68. Goecke IA, Bonilla 5, Marusic ET, Alvo M: EnhancedInsulin sensitivity In extrarenal potassium handling Inuremic rats. Kidney Int 199 1:39:39-43.

69. DeFronzo RA. Sherwin RS, Diffingham M: Influence ofbasal insulin and glucagon secretion on potassium andsodium metabolism: Studies with somatostatin in nor-mal dogs and In normal and diabetic human beings. JClin Invest 1978:61:472-479.

70. Allon M, Takeshian A, Shanklin N: Effect of Insulin-plus-glucose Infusion with or without epinephrine onfasting hyperkalemia. Kidney Int 1993:43:212-217.

71. Tzanialoukas AH, Avasthi PS: Temporal profile of serumpotassium concentration In nondiabetic and diabeticoutpatients on chronic dialysis. Am J Nephrol 1987:7:101-109.

72. Armstrong VW, Creutzfehdt W. Ebert R, Fuchs C, Hilg-ers R, Scheler F: Effect of dialysate glucose load onplasma glucose and glucoregulatory hormones In CAPDpatients. Nephron 1985:39:141-145.

73. Wideroe TE, Smeby LC. Myking OL, Wessel-Aas T:Glucose, insulin, and C-peptlde kinetics during contin-uous ambulatory peritoneal dialysis. Proc EDTA 1983:20:195-200.

74. Arrizabalaga P. Montoilu J, Martinez Vea A, Andreu L,Lopez Pedret J, Revert L: Increase in serum potassiumcaused by beta-2 adrenergic blockade In terminal renalfailure: Absence of mediation by insulin or aldosterone.Proc Eur Dial Transplant Assoc 1983:20:572-576.

75. Castellino P. Bia M, DeFronzo RA: Adrenergic modula-tion of potassium metabolism in uremia. Kidney Int1990:37:793-798.

76. Castro MCM, Freitas IF, Marcondes M, Sabbaga E: Pro-longed beta adrenergic stimulation-A new way to re-duce serum potassium concentration In hemodlalysis[Abstract]. Kidney Int 1994:46:1738.

77. Field MJ, Glebisch GJ: Hormonal control of renal potas-sium excretion. Kidney mt 1985:27:379-387.

78. Stanton B, Pan L, Deetjen H, Guckian V. Giebisch G:Independent effects of aldosterone and potassium oninduction of potassium adaptation In rat kidney. J ChinInvest 1987:79:198-206.

79. Sugarman A, Brown RS: The role of aldosterone inpotassium tolerance: Studies in anephrlc patients. Kid-ney Int 1988:34:397-403.

hyperkalemia in esrd

Documents