characteristics, risk factors, and adverse outcomes of...

Research ArticleCharacteristics Risk Factors and Adverse Outcomes ofHyperkalemia in Acute-on-Chronic Liver Failure Patients

Jun-jun Cai 1 KaiWang23 Hui-qing Jiang 1 and Tao Han 4

1Department of Gastroenterology e Second Hospital of Hebei Medical University Hebei Key Laboratory of GastroenterologyHebei Institute of Gastroenterology Shijiazhuang China

2Dongzhimen Hospital Beijing University of Chinese Medicine Beijing China3Sunsimiao Hospital Beijing University of Chinese Medicine Tongchuan China4Department of Hepatology Tianjinird Central Hospital Tianjin Institute of Hepatobiliary DiseaseTianjin Key Laboratory of Artificial Cell Tianjin China

Correspondence should be addressed to Hui-qing Jiang jianghuiqing1959sinacom and Tao Han hantaomd126com

Received 28 August 2018 Accepted 24 December 2018 Published 27 February 2019

Academic Editor Hongqun Liu

Copyright copy 2019 Jun-jun Cai et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Background Hyperkalemia is a serious complication in cirrhotic patients However the clinical characteristics risk factors and itsimpact on the outcomes in acute-on-chronic liver failure (ACLF) patients remain unclearMethods We retrospectively recruited650 ACLF patients in this study The risk factors associated with hyperkalemia and its relationship with 90-day mortality wereanalyzed using multivariable regression models Results Among 650 patients with ACLF 122 (79650) had hyperkalemia duringhospitalization Higher admission serum potassium levels and the presence of acute kidney injury (AKI) were independentrisk factors for hyperkalemia The prevalence rates of hyperkalemia in patients with and without AKI were 236 and 46respectively (Plt0001) Hyperkalemia was a predictor of mortality in AKI and non-AKI patients The 90-day mortality rates innon-AKI patients with and without hyperkalemia were 444 and 247 respectively (Plt0001) and in AKI patients with andwithout hyperkalemia were 803 and 566 respectively (Plt0001) Hepatic encephalopathy (HE) gastrointestinal bleeding AKIhyperkalemia elevated total bilirubin (TBIL) and international normalized ratio (INR) values and higher Model for End-StageLiver Disease (MELD) and chronic liver failure-sequential organ failure assessment (CLIF-SOFA) scores were independent riskfactors for predicting the 90-day mortality in ACLF patients Conclusions Hyperkalemia increases the 90-day mortality in ACLFpatients hyperkalemia is associated with AKI Patients with both AKI and hyperkalemia had the worst outcome

1 Introduction

Impaired potassium homoeostasis represents one of thecommonest electrolyte disturbances Both hypokalemia andhyperkalemia may have immediate deleterious physiologicaleffects and are consistently associated with adverse outcomes[1] The risk relationship between potassium levels andadverse outcomes is U-shaped with the lowest risk at serumpotassium of 4-45mmolL [2 3] Patients with advancedcirrhosis frequently have impaired potassium homoeostasisThe prevalence of hyperkalemia in this group of patients is12-14 while that in the general population is 21-70[4ndash6] Hyperkalemia may lead to arrhythmia inhibiting thecontraction and ventricular fibrillation and even death [7]

Recently a prospective multicenter study demonstrated thatall cirrhotic patients with a serumpotassiumconcentration ofge48mmolL (maximum 58mmolL) died within 1 year [8]

Acute-on-chronic liver failure (ACLF) is a devastatingentity characterized by acute deterioration of liver functionand one or more extrahepatic organ failures especiallykidney failure [9 10] The kidney plays a major role inpotassium homeostasis with kidney impairment being anespecially prominent risk factor for hyperkalemia A recentstudy showed that the incidence of hyperkalemia is 65higher among children with acute kidney injury (AKI)compared with those without AKI upon admission [11]Abnormal intake and excretionloss of potassium resultingfrom impairment of renal functionmay lead to hyperkalemia

HindawiBioMed Research InternationalVolume 2019 Article ID 6025726 9 pageshttpsdoiorg10115520196025726

2 BioMed Research International

923 patients with ACLF

273 patients were excluded for the following reasons 176 had hepatocellular carcinoma39 had server chronic extra-hepatic disease4 had human immunodeficiency virus infection2 received liver or kidney transplantation34 stay in hospital less than 48 hours18 aged outside 18-70-year old

391 patients without AKI(mortality rate 256)

259 patients with AKI(mortality rate 622)

650 patients with ACLF(mortality rate 402)

373 patients without hyperkalemia(mortality rate247)

18 patients with hyperkalemia(mortality rate 444)

198 patients without hyperkalemia(mortality rate 566)


Figure 1 Flow diagram of the study groups selection process ACLF acute-on-chronic liver failure AKI acute kidney injury

Several hormones including aldosterone antidiuretic hor-mone are involved in potassium metabolism Other riskfactors of hyperkalemia in combination with renal impair-ment include diabetes cardiovascular disease and use ofpharmacologic agents that affect the potassium regulationsuch as diuretic therapy potassium supplements and therenin-angiotensin-aldosterone system (RAAS) blockers [12ndash14] However the prevalence risk factors and impact ofhyperkalemia on the outcome of ACLF patients with orwithout renal impairment remain poorly defined Thereforethe present studywas to determine the prevalence risk factorsof hyperkalemia and its impact on the outcome of ACLFpatients with or without renal impairment

2 Methods

21 Materials and Methods A flow chart of the patient selec-tion process was presented in Figure 1 We retrospectivelyrecruited 650 patients with ACLF who were hospitalized atTianjin Third Central Hospital between March 2011 and July2016 ACLF was defined according to the criteria of AsianPacific Association for the Study of the Liver (APASL) Inbrief acute deterioration of liver function manifested byjaundice (total bilirubin [TBIL]ge5mgdL orge85120583molL) andcoagulopathy (international normalized ratio [INR] of pro-thrombin time ge15 or prothrombin activity [PTA] le40])complicated within 4 weeks with ascites andor hepaticencephalopathy in a patient with cirrhosis [15] Exclusion cri-teria are as follows (1) patients with hepatocellular carcinomaand nonhepatic neoplasia (2) immunocompromised patientswith human immunodeficiency virus infection (3) patientswith severe chronic extrahepatic disease (4) patients with a

hospital stay oflt 48 hours and (5) patients aged outside 18-70years old

Acute kidney injury (AKI) is defined as an increase inserum creatinine value by 03mgdL (265120583molL) within48 h or percentage by ge50 from baseline within 1 weekaccording to The Kidney Disease Improving Global Out-comes (KIDGO) criteria [16] Hyperkalemia was definedas at least one in-hospital serum potassium level measure-ment gt55mmolL The Model for End-Stage Liver Disease(MELD) score was calculated using the Malinchoc formula[17] The chronic liver failure-sequential organ failure assess-ment (CLIF-SOFA) score was calculated according to thecriteria of the European Association for the Study of theLiver-Chronic Liver Failure (EASL-CLIF) Consortium [9]

All patients were admitted and received standard sup-portive treatment Once recruited data were collected fromthe medical records including patient demographics vitalsigns complete blood count liver and renal function INRelectrolytes and complications of cirrhosis None of thepatients underwent liver transplantation and none of themwere lost to follow-up within the 90-day follow-up period

22 Ethics Statement The study performed in accordancewith the principle of the Declaration of Helsinki (2013 revi-sion) and approved by the Institutional Ethical Committee ofTianjin Third Central Hospital Written informed consentswere obtained during the hospital stay This trial was regis-tered in theChinese clinical trial registry ChiCTR1800017991(httpwwwchictrorgcnsearchchprojaspx)

23 Statistical Analysis Normally distributed variables wereexpressed as mean plusmn SD and nonnormally distributed

BioMed Research International 3

seru

m cr

eatin

ine (

umol

L)

serum potassium (mmolL)

r = 0201 lt

(a)


seru

m B

UN

(mm

olL

) lt

r = 021

(b)


lt

seru

m so

dium

(mm

olL

) r = -177

(c)

Figure 2 Correlations between serum potassium and serum creatinine serum BUN and serum sodium in patients with acute-on-chronicliver failure

variables were expressed as a median and interquartilerange (IQR) Count and percentages were used to describecategorical variables Two independent groups were com-pared using the t-test for continuous normally distributedvariables and the Mann-Whitney 119880 test for nonnormallydistributed variables For categorical variables comparisonsamong groups were made using the Chi-squared tests orFisherrsquos exact test as appropriate The Kaplan-Meier methodwas used to calculate the 90-day survival probability curveswhichwere comparedwith the log-rank test [18]Multivariatelogistic regression analyses were fitted to select the mainfactors associated with the endpoint Those found to have aP value of less than 005 on univariate analysis were includedin the multiple logistic regression models to evaluate theimpact on the presence of hyperkalemia and the 90-daymortality of patients Two-sided 119875 values of lt 005 wereconsidered statistically significant All statistical analyseswere performed using SPSS software version 170 (SPSS IncChicago IL USA)

3 Results

31 Characteristics of Patients Among 650 ACLF patients79 (122) presented hyperkalemia during hospitalizationThe characteristics of patientswith andwithout hyperkalemiaare summarized in Table 1 Patients with hyperkalemiahave a higher baseline serum potassium levels 47 (37-55) mmolL than those without hyperkalemia 39(35-43)mmolL (Plt0001) Hyperkalemia occurred more frequentlyin older patients with a higher white blood cell (WBC) countserum creatinine level BUN MELD score and CLIF-SOFAscore (Plt0001) and lower mean arterial pressure serumsodium and albumin levels (Plt005) These patients alsohad higher incidence of hepatic encephalopathy bacterialinfections acute kidney injury and gastrointestinal bleed-ing Interestingly no significant differences in the use ofpotassium-containing preparations mainly spironolactonepotassium chloride and potassium citrate were observedamong patients with and without hyperkalemia (Pgt005)

32 Correlation of Serum Potassium with Parameters of RenalFunction A positive correlation between serum potassiumand both serum creatinine (r=0201 Plt0001) and serum

BUN (r=0281 Plt0001) was detected in ACLF patients Bycontrast serum potassium had a negative correlation withserum sodium (r= -0177 Plt0001) (Figure 2)

33 Independent Risk Factors Associated with the Pres-ence of Hyperkalemia in ACLF Patients Multivariate logis-tic regression analysis showed that independent factors ofhyperkalemia were baseline serum potassium levels (oddsratio [OR]=348 95confidence interval[CI]=2309-5235Plt0001) and the presence of acute kidney injury (OR=47795CI= 2479-9166 and Plt0001) (Table 2)

34 Relationship between Hyperkalemia and Acute KidneyInjury in ACLF Patients Hyperkalemia was more prevalentin AKI patients (236) than in non-AKI patients (46)(Plt0001) No significant differences in age use of medica-tions etiology and complications of cirrhosis WBC ALBINR and parameters of liver function were observed amongAKI patients with and without hyperkalemia(Pgt005) Ofnote AKI patients with hyperkalemia had higher serumpotassium values and lower TBIL values than AKI patientswithout hyperkalemia (Plt005) (Table 3)

35 Effects of Hyperkalemia on 90-Day Survival in ACLFPatients Incidence of 90-day mortality in patients with andwithout hyperkalemia was 722 and 357 respectively(Plt0001) (Figure 3(a)) In patients with hyperkalemia non-survivors had a higher incidence of hepatic encephalopathyand acute kidney injury than survivors (Plt0001) Moreovernonsurvivors had a higher INR values serum potassiumconcentrations and MELD scores than survivors (Plt005)(Table 4)

In patients with hyperkalemia a multivariate logisticregression analysis was used to evaluate independent factorsassociated with 90-day mortality Based on the results ofthe univariate analysis of parameters presented in Table 4the independent factors associated with 90-day mortalitywere high INR values (OR=819 95CI=1668-40202 andP=0010) and acute kidney injury (OR=1136 95CI=1838-70186 and Plt 0009) (Table 5(a))

Due to the close association between hyperkalemia andacute kidney injury we further stratified the 90-daymortalityof patients with hyperkalemia according to the presence


Table 1 Characteristics of all patients according to the presence of hyperkalemia

variables without hyperkalemia With hyperkalemia P-valueN=571 N=79

Age (years) 510 plusmn 117 537 plusmn 110 0028Male-n 436(764) 65(823) 0241Diabetes-n 104(182) 15(190) 0868Hypertension-n 74(130) 13(165) 0392Etiology of cirrhosis-n ()Hepatitis B 327(573) 34(430) 0017Alcoholic 155(271) 35(443) 0002Hepatitis B plus alcoholic 30(53) 5(63) 0896Others 59(103) 5(63) 0263Medications-nDiuretic(frusemidespironolactone)

493(863) 68(861) 0993

Potassium chloride 263(461) 33(418) 0473Potassium citrate granules 279(489) 34(430) 0332Beta-blockers 37(65) 5(63) 0959Complications ofcirrhosis-nAscites 473(828) 67(848) 0661Bacterial infection 458(802) 74(937) 0004Hepatic encephalopathy 173(303) 33(418) 0040Acute kidney injury 198(347) 61(772) lt0001Gastrointestinal bleeding 111(194) 28(354) 0001Clinical and laboratorydata at admissionMAP(mmHg) 896 plusmn 122 847 plusmn 154 0001Heart rate(beatsminute) 842 plusmn 147 850 plusmn 166 0658WBC(times109L) 77 plusmn 49 99 plusmn 66 lt0001PLT(times109L) 780(510-1235) 985(480-1300) 0589ALB(gL) 283 plusmn 52 269 plusmn 58 0026TBIL(120583molL ) 2169(1302-3212) 1442(1045-2997) 0128INR 21(17-26) 22(18-29) 0177BUN(mmolL) 54(40-89) 96(52-144) lt0001Serum Cr(120583molL) 610(490-840) 870(580-1220) lt0001Serum Na+ (mmolL) 1341(1303-1371) 1313(1251-1356) lt0001Serum K+ (mmolL) 39(34-43) 47(37-55) lt0001CTP score 120(100-130) 120(110-130) 0066MELD score 219 plusmn 71 251 plusmn 88 lt0001CLIF-SOFA score 70(60-80) 70(70-110) 0002Number of deaths-n 204(357) 57(722) lt0001MAP mean arterial pressure WBC white blood cells PLT platelet ALB albumin TBIL total bilirubin INR international normalized ratio BUN blooduria nitrogen Cr creatinine CTP Child-Turcotte-Pugh MELD Model for End-Stage Liver Disease CLIF-SOFA chronic liver failure-sequential organ failureassessment

Table 2 Risk factors for the presence of hyperkalemia in patients with acute-on-chronic liver failure

Effect Estimate OR(95CI) Standard error Wald X2 P-valueAcute kidney injury 1562 477(2479-9166) 0334 21907 lt0001Admission potassium levels (mmolL) 1246 348(2309-5235) 0209 35622 lt0001


Table 3 Characteristics of ACLF patients with acute kidney injury according to presence of hyperkalemia

VariablesNO AKI AKI

Plowast-valueNo hyperkalemia Hyperkalemia No hyperkalemia HyperkalemiaN=373 N=18 N=198 N=61

Age (years) 501 plusmn 117 501 plusmn 127 528 plusmn 115 548 plusmn 103 0115Male-n 281(753) 15(833) 155(783) 50(820) 1000Diabetes-n 63(169) 3(167) 39(197) 11(180) 1000Hypertension-n 45(121) 4(222) 30(152) 9(148) 0479Etiology of cirrhosis-n ()Hepatitis B 87(233) 7(188) 83(419) 28(459) 0599Alcoholic 231(619) 11(29) 80(404) 23(377) 0078Hepatitis B plus alcoholic 23(62) 0 9(45) 5(82) 0583Others 32(86) 0 26(131) 5(82) 0583Medications-nDiuretic (frusemidespironolactone) 309(828) 14(778) 183(924) 55(902) 0324

Potassium chloride 155(416) 5(278) 108(545) 28(459) 0171Potassium citrate granules 184(493) 7(389) 95(480) 27(443) 0686Beta-blockers 22(59) 0 15(76) 5(82) 0583Complications ofcirrhosis-nAscites 300(804) 13(722) 173(874) 54(885) 0090Bacterial infection 273(732) 15(833) 185(934) 59(967) 0075Hepatic encephalopathy 93(249) 4(222) 79(399) 29(475) 0063Gastrointestinal bleeding 53(142) 3(167) 58(293) 25(410) 0091Clinical and laboratorydata at admissionMAP(mmHg) 823 plusmn 146 829 plusmn 112◼ 868 plusmn 139 852 plusmn 165 0578Heart rate(beatsminute) 912 plusmn 110 811 plusmn 194 877 plusmn 144 862 plusmn 157 0269WBC(times109L) 68 plusmn 40 65 plusmn 41 94 plusmn 58 109 plusmn 68 0010PLT(times109L) 810(510-1260) 755(318-1513) 725(498-1150) 900(550-1260) 0444ALB(gL) 291 plusmn 52 285 plusmn 55 268 plusmn 50 264 plusmn 58 0177TBIL(120583molL ) 2104(1313-2907) 2815(1067-3436) 2255(1172-3609) 1686(1042-2683)998771 0086INR 20(17-25) 23(18-34) 22(17-29) 22(18-28) 0532BUN(mmolL) 48(36-65) 52(36-84) 89(51-142) 105(60-162) 0002Serum Cr(120583molL) 550(460-670) 635(538-903)◼ 880(630-1390) 950(675-1280) 0013Serum Na+ (mmolL) 1347(1317-1375) 1328(1350-1370) 1323(1279-1356) 1304(1250-1353) 0516Serum K+(mmolL) 39(34-42) 42(36-56)◼ 39(34-43) 48(38-55)998771 0599CTP score 110(100-120) 120(100-130) 120(110-130) 120(110-130) 0398MELD score 198 plusmn 57 245 plusmn 76◼ 259 plusmn 80 252 plusmn 91 0767CLIF-SOFA score 70(60-80) 75(70-113)◼ 80(70-100) 70(65-110) 0275Plowast value between patients with and without AKI with hyperkalemia P◼lt005 patients without AKI no hyperkalemia versus hyperkalemia P998771lt005patients with AKI no hyperkalemia versus hyperkalemiaMAP mean arterial pressure WBC white blood cells PLT platelet ALB albumin TBIL total bilirubin INR international normalized ratio BUN blood ureanitrogen Cr creatinine CTP Child-Tturcotte-Pugh MELD Model for End-Stage Liver Disease CLIF-SOFA chronic liver failure-sequential organ failureassessment

of acute kidney injury The corresponding survival curvesof AKI and non-AKI patients according to incident hyper-kalemia are shown in Figure 3(b) The 90-day mortalityrates of non-AKI patients with and without hyperkalemiawere 444 and 247 respectively (Plt0001) By con-trast the 90-day mortality rates of AKI patients with and

without hyperkalemia were 803 and 566 respectively(Plt0001)

A multivariate logistic regression analysis of all ACLFpatients showed that HE gastrointestinal bleeding AKI andhyperkalemia elevated TBIL and INR values high MELDand CLIF-SOFA scores were independent risk factors for


571 367384442

79 222541

No hyperkalemia

Hyperkalemia

Log-rank test Plt0001

30 60 900Time (days)

prob

abili

ty of s

urviva

l (

)No hyperkalemia

Hyperkalemia

(a)

373 28129532518 101113

198 868911761 121428


no AKI without hyperkalemiano AKI with hyperkalemiaAKI without hyperkalemiaAKI with hyperkalemia

no AKI without hyperkalemia

no AKI with hyperkalemia

AKI without hyperkalemia

AKI with hyperkalemia

prob

abili

ty of s

urviva

l (

)


(b)

Figure 3 Kaplan-Meier survival curves (a) in patientswith andwithout hyperkalemia (b) in patientswith hyperkalemia according to presenceof acute kidney injury (AKI)

predicting the 90-day mortality with ORs of 334 200 235298 100 177 090 and 119 respectively (Table 5(b))

4 Discussion

Hyperkalemia is one of the most serious electrolyte distur-bances with an associated sudden cardiac death and fatalarrhythmias Under normal conditions serum potassiumconcentration is strictly regulated by redundancy of home-ostatic mechanism and maintained within a narrow rangeof 35-55mmolL Once these homeostatic mechanisms aredisrupted potassium abnormalities may occur With theincreased in serum potassium concentrations the risk ofadverse outcome increases substantially whichmakes hyper-kalemia a medical emergency that need special attentionHowever in clinical practice no single threshold couldidentify the imminent adverse outcome of hyperkalemia indifferent patients [19]

In our study 122 ACLF patients developed hyper-kalemia during hospitalization Hyperkalemia commonlyaffects older patients with a higher WBC count moresevere hepatorenal function and higher incidence of hepaticencephalopathy bacterial infections gastrointestinal bleed-ing and AKI Patients with advanced cirrhosis manifesta hyperdynamic circulatory state which is a progressivevasodilatory syndrome and a key factor in the pathogenesis ofvarious complications of cirrhosis prominently in the kidney[20] Vasodilatation leads to a decreased effective arterialblood volume and activation of neurohumoral systems suchas RAAS Increased plasma antidiuretic hormone (ADH)levels lead to water and sodium retention causing ascitesand dilution hyponatremia further [4] In hyponatremiathe Na+-K+-ATPase pumps sodium ions out of the celland potassium into the cell [21] In addition patients withcirrhosis are characterized by a state of secondary hyperal-dosteronism Aldosterone a mineralocorticoid hormone is

known to regulate the sodium reabsorption and potassiumsecretion in renal cortical-collecting duct Aldosterone actsby increasing the number of open sodium channels leadingto increased sodium reabsorption and potassium secretionThis mechanism explains the negative relationship betweenserum potassium and serum sodium in ACLF patients

Gastrointestinal bleeding increased diuretic doses largevolume paracentesis and lactulose-induced diarrhea maylead to a state of hypovolemia inACLF patients Hypovolemicactivates the renin-angiotensin system [21] The increasedcirculating aldosterone stimulates renal sodium retentionand affects potassium excretion This perturbed potassiumhomoeostasis is further disturbed by pharmacologic agentssuch as antialdosterones and inhibitors of RAAS which areknown to affect serum potassium Therefore hyperkalemiacan be considered a late event in the natural history ofcirrhosis which occurs after the onset of ascites gastroin-testinal bleeding and renal impairment [4]Of note diureticssuch as spironolactone and various potassium-containingpreparations (potassium chloride potassium citrate andpotassium magnesium aspartate) are commonly used in themanagement of cirrhotic patients and may lead to hyper-kalemia However in our study no significant differencesin the use of potassium-saving drugs were observed amongpatients with and without hyperkalemia (Pgt005)

We also found that higher baseline serum potassiumlevels and the presence of AKI were the independent riskfactors of hyperkalemia We further analyzed the relationshipbetween hyperkalemia and AKIWe found that hyperkalemiawas more frequently seen in patients with AKI Surprisinglyno significant differences in age etiology and complicationsof cirrhosis and other parameters except for the higher base-line serum potassium concentration were observed amongAKI patients with and without hyperkalemia on admissionHence it is difficult for clinicians to identify the early pres-ence of hyperkalemia in AKI patients However the 90-day


Table 4 Comparison between patients with hyperkalemia who died or survived

Variables Survival Non-survival P-valueN=22 N=57

Age (years) 525 plusmn 79 542 plusmn 120 0563Male-n 17(773) 48(842) 0469Etiology of cirrhosis-n ()Hepatitis B 7(318) 23(404) 0484Alcoholic 12(545) 26(456) 0476Hepatitis B plus alcoholic 1(45) 4(70) 1000others 2(91) 3(53) 0614Medications-nDiuretic (frusemidespironolactone) 20(909) 48(842) 0683Potassium chloride 11(500) 22(386) 0357Potassium citrate granules 9(409) 25(439) 0812Beta-blockers 1(45) 4(70) 1000Complications of cirrhosis-nAscites 17(773) 50(877) 0246Bacterial infection 20(909) 54(947) 0614Hepatic encephalopathy 5(227) 28(491) 0033Acute kidney injury 12(545) 49(860) 0003Gastrointestinal bleeding 5(227) 23(404) 0142Clinical and laboratory data at admissionMAP(mmHg) 626 plusmn 157 837 plusmn 159 0327Heart rate(beatsminute) 818 plusmn 117 863 plusmn 180 0292WBC(times109L) 85 plusmn 48 105 plusmn 71 0248PLT(times109L) 1040(725-152) 770(470-1205) 0056ALB(gL) 284 plusmn 37 263 plusmn 63 0151TBIL(120583molL ) 2288(1044-3436) 1843(1066-2869) 0314INR 18(16-21) 25(18-33) 0003BUN(mmolL) 66(43-108) 105(56-153) 0057Serum Cr(120583molL) 706(540-1040) 950(600-1290) 0077Serum Na+ (mmolL) 1322(1246-1371) 1310(1252-1356) 0600Serum K+ (mmolL) 38(34-55) 49(40-56) 0038CTP score 115(100-130) 120(110-130) 0079MELD score 221 plusmn 44 262 plusmn 97 0016CLIF-SOFA score 70(70-100) 80(70-115) 0596MAP mean arterial pressure WBC white blood cells PLT platelet ALB albumin TBIL total bilirubin INR international normalized ratio BUN bloodurea nitrogen Cr creatinine CTP Child-Turcotte-Pugh MELD Model for End-Stage Liver Disease CLIF-SOFA chronic liver failure-sequential organ failureassessment

mortality was significantly higher in AKI patients with hyper-kalemia than in non-AKI patients without hyperkalemiaMoreover hyperkalemia is the independent risk factor forthe 90-daymortality of ACLF patients The physiopathologicassociation between the hyperkalemia and poor prognosis iscomplex and remains to be elucidated The mechanisms thatdrive this poor prognosis are likely due to the occurrence of asevere disease and the use of therapeutic drugs which in caseswith a renal dysfunction may result in hyperkalemia [22]This result is in line with our findings indicating the positiverelationship between serum potassium and creatinine as wellas BUN which are known markers of renal dysfunctionHence it reminds us that appropriate surveillance regimensand the timely correction of abnormal serum potassium

levels and renal function are critical especially in AKIpatients

This study had limitations Firstly this was a retrospectiveand single center study the results need to be validatedin multicenter prospective study Secondly this study mayunderestimate the incidence and risk of potassium varia-tions in clinical practice where laboratory surveillance isfrequently less assiduous and regular

In conclusion 122 of ACLF patients had hyperkalemiaand a close association between hyperkalemia and AKI wasdetected in ACLF patients Hyperkalemia is more frequentlyseen in patients with AKI The 90-day mortality rates wereclearly different among AKI patients with and withouthyperkalemia AKI patients with hyperkalemia had the worst


Table 5 Independent factors associated with 90-day mortality (a) in patients with hyperkalemia and (b) in all ACLF patients

(a)

Effect Estimate OR (95CI) Standard error Wald X2 P-valueAcute kidney injury 243 1136(1838-70186) 0929 6837 0009INR 2103 819(1668-40202) 0812 6707 0010

(b)

Effect Estimate OR (95CI) Standard error Wald X2 P-valueHepatic encephalopathy 1206 334(2146-5195) 0226 28560 lt0001Gastrointestinal bleeding 0693 200(1207-3315) 0258 7231 0007Acute kidney injury 0853 235(1482-3714) 0234 13241 lt0001Hyperkalemia 1091 298(1539-5760) 0337 10505 0001TBIL 0004 100(1002-1006) 0001 13893 lt0001INR 0568 177(1230-2532) 0184 9526 0002MELD -0110 090(0826-0973) 0042 6888 0009CLIF-SOFA 0176 119(1065-1335) 0058 9307 0002INR international normalized ratio TBIL total bilirubin MELD Model for End-Stage Liver Disease CLIF-SOFA chronic liver failure-sequential organfailure assessment

outcome HE gastrointestinal bleeding AKI hyperkalemiaelevated TBIL and INR values and high MELD and CLIF-SOFA scores were independent factors for predicting the 90-day mortality in ACLF patients The dynamic monitoringand timely treatment of abnormal serum potassium levelsand renal function are critical for improving the patientsrsquoprognosis

Abbreviations

ACLF Acute-on-chronic liver failureAKI Acute kidney injuryCTP Child-turcotte-pughMELD Model for End-Stage Liver DiseaseCLIF-SOFA Chronic liver failure-sequential organ

failure assessmentHE Hepatic encephalopathyWBC White blood cellsPLT PlateletALB AlbuminTBIL Total bilirubinBUN Blood urea nitrogenCr CreatinineINR International normalized ratio

Data Availability

All data are available on the website of Chinese clinical trialregistry within six months after the manuscript published

Conflicts of Interest

The authors declared no conflicts of interest

Acknowledgments

This work was supported by the National 13th 5-year Planfor Hepatitis Research [Grant no 2017ZX10203201-007] and

Graduate Innovation Fund of Education Department ofHebei Province [Grant no CXZZBS2017104]

References

[1] A S Desai J Liu M A Pfeffer et al ldquoIncident HyperkalemiaHypokalemia and Clinical Outcomes During SpironolactoneTreatment of Heart Failure With Preserved Ejection FractionAnalysis of the TOPCAT Trialrdquo Journal of Cardiac Failure vol24 no 5 pp 313ndash320 2018

[2] C P Kovesdy K Matsushita Y Sang et al ldquoSerum potassiumand adverse outcomes across the range of kidney function aCKD Prognosis Consortium meta-analysisrdquo European HeartJournal vol 39 no 17 pp 1535ndash1542 2018

[3] J Nunez A Bayes-Genıs F Zannad et al ldquoLong-Term Potas-sium Monitoring and Dynamics in Heart Failure and Risk ofMortalityrdquo Circulation vol 137 no 13 pp 1320ndash1330 2018

[4] R Maiwall S Kumar M K Sharma Z Wani M Ozukumand S K Sarin ldquoPrevalence and prognostic significance ofhyperkalemia in hospitalized patients with cirrhosisrdquo Journalof Gastroenterology and Hepatology vol 31 no 5 pp 988ndash9942016

[5] E Nilsson A Gasparini J Arnlov et al ldquoIncidence anddeterminants of hyperkalemia and hypokalemia in a largehealthcare systemrdquo International Journal of Cardiology vol 245pp 277ndash284 2017

[6] Y Chen A R Chang M A McAdams DeMarco et al ldquoSerumPotassium Mortality and Kidney Outcomes in the Atheroscle-rosis Risk in Communities StudyrdquoMayo Clinic Proceedings vol91 no 10 pp 1403ndash1412 2016

[7] A Goyal J A Spertus K Gosch et al ldquoSerum potassium levelsand mortality in acute myocardial infarctionrdquo Journal of theAmerican Medical Association vol 307 no 2 pp 157ndash164 2012

[8] S Wallerstedt M Simren S Wahlin et al ldquoModerate hyper-kalemia in hospitalized patientswith cirrhotic ascites indicates apoor prognosisrdquo Scandinavian Journal of Gastroenterology vol48 no 3 pp 358ndash365 2013

[9] R Moreau R Jalan P Gines et al ldquoAcute-on-chronic liverfailure is a distinct syndrome that develops in patients with


acute decompensation of cirrhosisrdquo Gastroenterology vol 144no 7 pp 1426e9ndash1437e9 2013

[10] P Angeli E Rodrıguez S Piano et al ldquoAcute kidney injuryand acute-on-chronic liver failure classifications in prognosisassessment of patients with acute decompensation of cirrhosisrdquoGut vol 64 no 10 pp 1616ndash1622 2015

[11] F Chowdhury P K Ghosh K M Shahunja et al ldquoHyper-kalemia Was an Independent Risk Factor for Death WhileUnder Mechanical Ventilation Among Children HospitalizedWith Diarrhea in Bangladeshrdquo Global Pediatric Health vol 5Article ID 2333794X1775400 2018

[12] M Legrand P-O Ludes Z Massy et al ldquoAssociation betweenhypo- and hyperkalemia and outcome in acute heart failurepatients the role of medicationsrdquo Clinical Research in Cardiol-ogy vol 107 no 3 pp 214ndash221 2018

[13] G Bandak Y Sang A Gasparini et al ldquoHyperkalemia afterinitiating renin-angiotensin system blockade The StockholmCreatinine Measurements (SCREAM) projectrdquo Journal of theAmerican Heart Association vol 6 no 7 2017

[14] P H Pun B A Goldstein J A Gallis J P Middleton andL P Svetkey ldquoSerum Potassium Levels and Risk of SuddenCardiac Death Among Patients With Chronic Kidney Diseaseand Significant Coronary Artery Diseaserdquo Kidney InternationalReports vol 2 no 6 pp 1122ndash1131 2017

[15] S K Sarin C K Kedarisetty Z Abbas et al ldquoAcute-on-chronic liver failure consensus recommendations of the AsianPacific Association for the Study of the Liver (APASL) 2014rdquoHepatology International vol 8 no 4 pp 453ndash471 2014

[16] A K I Kdigo Kidney Disease Improving Global Outcomes vol1 AKIWorkGroup Clinical practice guideline for acute kidneyinjury 2012

[17] A E Ruf W K Kremers L L Chavez V I Descalzi L GPodesta and F G Villamil ldquoAddition of serum sodium into theMELD score predicts waiting list mortality better than MELDalonerdquo Liver Transplantation vol 11 no 3 pp 336ndash343 2005

[18] J Cai K Wang T Han and H Jiang ldquoEvaluation of prognosticvalues of inflammation-based makers in patients with HBV-related acute-on-chronic liver failurerdquoMedicine vol 97 no 46Article ID e13324 2018

[19] R PM TeDorsthorst J HendrikseM TVervoorn V YH vanWeperen andMA G van der Heyden ldquoReview of case reportson hyperkalemia induced by dietary intake not restricted tochronic kidney disease patientsrdquo European Journal of ClinicalNutrition pp 1ndash8 2018

[20] Y Iwakiri and R J Groszmann ldquoThe hyperdynamic circulationof chronic liver diseases From the patient to the moleculerdquoHepatology vol 43 no 2 pp S121ndashS131 2006

[21] U Udensi and P Tchounwou ldquoPotassium homeostasis oxida-tive stress and human diseaserdquo International Journal of Clinicaland Experimental Physiology vol 4 no 3 pp 111ndash122 2017

[22] S S Sheen R W Park D Yoon G-T Shin H Kim andI-W Park ldquoThe Model for End-stage Liver Disease scoreis potentially a useful predictor of hyperkalemia occurrenceamong hospitalized angiotensin receptor blocker usersrdquo Journalof Clinical Pharmacy anderapeutics vol 40 no 1 pp 48ndash542015

Stem Cells International

Hindawiwwwhindawicom Volume 2018


MEDIATORSINFLAMMATION

of

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018

Immunology ResearchHindawiwwwhindawicom Volume 2018

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine


Behavioural Neurology

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinsonrsquos Disease

Evidence-Based Complementary andAlternative Medicine

Volume 2018Hindawiwwwhindawicom

Submit your manuscripts atwwwhindawicom


923 patients with ACLF

273 patients were excluded for the following reasons 176 had hepatocellular carcinoma39 had server chronic extra-hepatic disease4 had human immunodeficiency virus infection2 received liver or kidney transplantation34 stay in hospital less than 48 hours18 aged outside 18-70-year old

391 patients without AKI(mortality rate 256)

259 patients with AKI(mortality rate 622)

650 patients with ACLF(mortality rate 402)

373 patients without hyperkalemia(mortality rate247)

18 patients with hyperkalemia(mortality rate 444)



Figure 1 Flow diagram of the study groups selection process ACLF acute-on-chronic liver failure AKI acute kidney injury

Several hormones including aldosterone antidiuretic hor-mone are involved in potassium metabolism Other riskfactors of hyperkalemia in combination with renal impair-ment include diabetes cardiovascular disease and use ofpharmacologic agents that affect the potassium regulationsuch as diuretic therapy potassium supplements and therenin-angiotensin-aldosterone system (RAAS) blockers [12ndash14] However the prevalence risk factors and impact ofhyperkalemia on the outcome of ACLF patients with orwithout renal impairment remain poorly defined Thereforethe present studywas to determine the prevalence risk factorsof hyperkalemia and its impact on the outcome of ACLFpatients with or without renal impairment

2 Methods

21 Materials and Methods A flow chart of the patient selec-tion process was presented in Figure 1 We retrospectivelyrecruited 650 patients with ACLF who were hospitalized atTianjin Third Central Hospital between March 2011 and July2016 ACLF was defined according to the criteria of AsianPacific Association for the Study of the Liver (APASL) Inbrief acute deterioration of liver function manifested byjaundice (total bilirubin [TBIL]ge5mgdL orge85120583molL) andcoagulopathy (international normalized ratio [INR] of pro-thrombin time ge15 or prothrombin activity [PTA] le40])complicated within 4 weeks with ascites andor hepaticencephalopathy in a patient with cirrhosis [15] Exclusion cri-teria are as follows (1) patients with hepatocellular carcinomaand nonhepatic neoplasia (2) immunocompromised patientswith human immunodeficiency virus infection (3) patientswith severe chronic extrahepatic disease (4) patients with a

hospital stay oflt 48 hours and (5) patients aged outside 18-70years old

Acute kidney injury (AKI) is defined as an increase inserum creatinine value by 03mgdL (265120583molL) within48 h or percentage by ge50 from baseline within 1 weekaccording to The Kidney Disease Improving Global Out-comes (KIDGO) criteria [16] Hyperkalemia was definedas at least one in-hospital serum potassium level measure-ment gt55mmolL The Model for End-Stage Liver Disease(MELD) score was calculated using the Malinchoc formula[17] The chronic liver failure-sequential organ failure assess-ment (CLIF-SOFA) score was calculated according to thecriteria of the European Association for the Study of theLiver-Chronic Liver Failure (EASL-CLIF) Consortium [9]

All patients were admitted and received standard sup-portive treatment Once recruited data were collected fromthe medical records including patient demographics vitalsigns complete blood count liver and renal function INRelectrolytes and complications of cirrhosis None of thepatients underwent liver transplantation and none of themwere lost to follow-up within the 90-day follow-up period

22 Ethics Statement The study performed in accordancewith the principle of the Declaration of Helsinki (2013 revi-sion) and approved by the Institutional Ethical Committee ofTianjin Third Central Hospital Written informed consentswere obtained during the hospital stay This trial was regis-tered in theChinese clinical trial registry ChiCTR1800017991(httpwwwchictrorgcnsearchchprojaspx)

23 Statistical Analysis Normally distributed variables wereexpressed as mean plusmn SD and nonnormally distributed


seru

m cr

eatin

ine (

umol

L)


r = 0201 lt

(a)


seru

m B

UN

(mm

olL

) lt

r = 021

(b)


lt

seru

m so

dium

(mm

olL

) r = -177

(c)



3 Results













493(863) 68(861) 0993






VariablesNO AKI AKI








571 367384442

79 222541

No hyperkalemia

Hyperkalemia



prob

abili

ty of s

urviva

l (

)No hyperkalemia

Hyperkalemia

(a)

373 28129532518 101113

198 868911761 121428







prob

abili

ty of s

urviva

l (

)


(b)



4 Discussion
















(a)


(b)



Abbreviations



Data Availability




Acknowledgments



References





























of




Disease Markers



OncologyJournal of





PPAR Research



Volume 2018


Journal of

ObesityJournal of




















seru

m cr

eatin

ine (

umol

L)


r = 0201 lt

(a)


seru

m B

UN

(mm

olL

) lt

r = 021

(b)


lt

seru

m so

dium

(mm

olL

) r = -177

(c)



3 Results













493(863) 68(861) 0993






VariablesNO AKI AKI








571 367384442

79 222541

No hyperkalemia

Hyperkalemia



prob

abili

ty of s

urviva

l (

)No hyperkalemia

Hyperkalemia

(a)

373 28129532518 101113

198 868911761 121428







prob

abili

ty of s

urviva

l (

)


(b)



4 Discussion
















(a)


(b)



Abbreviations



Data Availability




Acknowledgments



References





























of




Disease Markers



OncologyJournal of





PPAR Research



Volume 2018


Journal of

ObesityJournal of























493(863) 68(861) 0993






VariablesNO AKI AKI








571 367384442

79 222541

No hyperkalemia

Hyperkalemia



prob

abili

ty of s

urviva

l (

)No hyperkalemia

Hyperkalemia

(a)

373 28129532518 101113

198 868911761 121428







prob

abili

ty of s

urviva

l (

)


(b)



4 Discussion
















(a)


(b)



Abbreviations



Data Availability




Acknowledgments



References





























of




Disease Markers



OncologyJournal of





PPAR Research



Volume 2018


Journal of

ObesityJournal of





















VariablesNO AKI AKI








571 367384442

79 222541

No hyperkalemia

Hyperkalemia



prob

abili

ty of s

urviva

l (

)No hyperkalemia

Hyperkalemia

(a)

373 28129532518 101113

198 868911761 121428







prob

abili

ty of s

urviva

l (

)


(b)



4 Discussion
















(a)


(b)



Abbreviations



Data Availability




Acknowledgments



References





























of




Disease Markers



OncologyJournal of





PPAR Research



Volume 2018


Journal of

ObesityJournal of




















571 367384442

79 222541

No hyperkalemia

Hyperkalemia



prob

abili

ty of s

urviva

l (

)No hyperkalemia

Hyperkalemia

(a)

373 28129532518 101113

198 868911761 121428







prob

abili

ty of s

urviva

l (

)


(b)



4 Discussion
















(a)


(b)



Abbreviations



Data Availability




Acknowledgments



References





























of




Disease Markers



OncologyJournal of





PPAR Research



Volume 2018


Journal of

ObesityJournal of





























(a)


(b)



Abbreviations



Data Availability




Acknowledgments



References





























of




Disease Markers



OncologyJournal of





PPAR Research



Volume 2018


Journal of

ObesityJournal of






































of




Disease Markers



OncologyJournal of





PPAR Research



Volume 2018


Journal of

ObesityJournal of



















characteristics, risk factors, and adverse outcomes of...

Documents