Prebypass glucose-insulin-potassium infusion in elective nondiabetic coronary artery surgery patients

ORIGINAL ARTICLES

Prebypass

Glucose-Insulin-Potassium Infusion in Elective Artery Surgery Patients Jan-Ola

Perioperative adequate However,

M. Wistbacka,

GIK therapy

MD, Päivi K. Kaukoranta,

has been advocated

to ensure

energy substrate levels during cardiac surgery. hyperglycemia should be avoided because it may

worsen neurologie outcome spective,

randomized,

between

two

after cerebral ischemia. A pro-

clinical comparison

prebypass

infusion

nondiabetic CABG patients. received glucose, 0.6 g/kg/h,

regimens

was performed in 32 elective

Sixteen patients (GIK group) insulin, 0.12 U/kg/h, and KCI,

0.12 mmol/kg/h, from the induction of anesthesia to the start of CPB; while the remaining 16 patients (R group) received only Ringer’s acetate. The pump prime was glucose free and a blood cardioplegia

technique

was used in both

groups. NO differences were found between the groups with regard to myocardial were

elevated

injury; the CK-MB enzyme

to a similar

degree

fractions

and the frequency

of

postoperative ECG changes were similar in both groups. Likewise, there were no differences in hemodynamic changes, need for inotropic support, arrhythmia frequency, or duration

S

URGICAL STRESS results in metabolic changes related to increased neuroendocrine activity,1,2 leading to increased lipolysis, elevated blood glucose levels3s4 impaired glucose tolerante and utilization,5 and peripheral insulin resistance.‘j Under aerobic conditions the heart predominantly uses free fatty acids (FFAs), but during ischemia there is a shift from oxidative lipolysis to anaerobit glycolysis.7 Myocardial FFA oxidation is limited during reperfusions and lipolysis and FFAs seem to be harmful during and after ischemia, with depressed myocardial function and an increased incidence of arrhythmias.9-i1 Perioperative glucose-insulin-potassium (GIK) therapy has been advocated during cardiac surgery to ensure adequate myocardial glycogen levels, enhance anaerobic glycolysis, and suppress lipolysis and fatty acid liberation.12J3 On the other hand, intravenous glucose infusion may enhance stress-induced hyperglycemia. During cardiac surgery there is always a potential risk of cerebral ischemia,14 and hyperglycemia should be avoided because it may worsen the outcome after ischemic cerebral events.15-r7 Hence, it is stil1 an open question whether or not the benefits of perioperative GIK infusion outweigh the possible disadvantages. This study was designed to evaluate two prebypass infusion regimens in elective coronary artery bypass graft (CABG) patients with respect to the hyperglycemic response, hemodynamic changes and inotropic support, electrocardiogram changes and arrhythmias, CK-MB enzyme changes, and duration of intensive care unit (ICU) stay.

Nondiabetic

MD, and Lauri S. Nuutinen,

Coronary

MD, PhD

of ICU stay. The GIK patients had higher blood glucose (P c 0.05) and insulin levels (P c 0.01); blood glucose increased to 12.4 + 5.4 mmol/L (mean + SD) at cannulation, with a drop after starting bypass. Interindividual variation in GIK patients was great, with glucose values ranging between 20.1 mmol/L at cannulation to 2.0 mmol/L after starting CPB. A hyperglycemic response was seen in both groups during rewarming: 15.0 + 4.2 and 15.0 f 3.1 mmol/L in GIK and R patients, respectively. It is concluded that prebypass GIK infusion had no clinical benefits for elective CABG patients as compared to Ringer’s acetate. Blood glucose rose in GIK patients before CPB, but in two patients it fell to a potentially dangerous hypoglycemie leve1 after starting bypass. Thus, GIK should be used with caution due to the potential risk of serious side effects. Copyright o 1992 by W.B. Saunders Company KEY WORDS: glucose, insulin, coronary surgery, cardiopulmonary bypass PATIENTS AND METHODS This prospective, randomized trial was approved by the Ethics Committee for Medical Research at Oulu University Hospita]; al1 of the patients gave their informed consent. The patients had NYHA class 11-IV symptoms, and their preoperative left ventricular ejection fraction was at least 50%. Patients with known endocrine or metabolic diseases were excluded. Al1 of the patients were on sublingual or long-acting nitrates, P-blockers, andior calcium Channel antagonists with a reduced dose included in their

This article is accompanied by an editorial. Wiese S, Askanazi J: Glucose/Insulin/Potassium monary

Bypass.

J Cardiothorac

Vast

Anesth

Please see: Therapy: 6:517-520,

From the Depattment of Anesthesia, Oulu University Hospital, Oulu, Finland. This work was supported by a gram frons Finska Läkaresällskapet, Helsinki, Finland. Presented in part at EACTA ‘90, 5th Annual Meeting, held in Vienna, Austria, May 1990, and the Finnish Society OfAnaesthesiologists’Annua1 Meeting, held in Helsinki, Finland, November 1990. Address reprint requests to Jan-Ola Wistbacka, MD, Depatiment of Anesthesia, Oulu University Hospital, PO Box 22, SF-90220 Oulu, Finland. Copyright 0 1992 by W.B. Saunders Company 1053-0770/9210605-0002$03.0010

Journalof Cardiofhoracicand VascularAnesthesia, Vol 6, NO5 (October), 1992: pp 521-527

521

522

WISTBACKA

premedication. Preoperative fasting was for 12 houra. Clinical details are given in Tables I and 2. Thirty-two well-nourished nondiabetic patients scheduled for elective CABG surgery were rdndomly allocated into two groups. The GIK group received glucose, 0.6 g/kg/h, rapid-acting insulin. 0.12 Uikgih, and KCI. 0.12 mmolikgih. Infusion into a centralvein was started after cannulation and preoperative hlood sample collection and was stopped immediately after the onset of cardiopulmonary bypass (CPB). Infusion time was 1.5 to 2 hours. The control group. group R, received only glucose-free Ringer’s acetate Premedication consisted of lorazepam, 0.03 mgikg, oraily in the evening, and diazepam, 0.15 mgikg, orally in the morning. A modified high-dose fentanyl anesthetic (0.03 mgikg + 0.01 mg/ kgih) with lorazepam. 0.03 mgikg, was used. Pancuronium, 0.10 mgikg, was given for intubation, and incremental doses of pancuronium and diazepam were administered before surgery and CPB. The patients were ventilated with 40% oxygen in air supplemented with enflurane if needed. A membrane oxygenator (Dideco Compactflo; Dideco S.p.A., Mirandola, Italy, or Shiley M2000; Shiley Corporation, Irvine, CA) was used with a nonpulsatile pump flow of 1.5 to 1.8Liminlm” with hypothermia of 28°C. The pump priming solutions were glucose free, consisting of 0.9% saline, 1,000 mL, Ringer’s acetate, 400 to 1,000 mL, and 15’% mannitol, 250 mL. A blood cardioplegia technique modified from Buckbergtk was used for cardiac protection by means of a Dideco D512 cardioplegia delivery set (Dideco S.p.A., Mirandola, Italy). Cardiac arrest was initiated with a warm (35 to 36°C) hyperkalemic solution (solution B 1 in Table 3) and oxygenated blood infused into the aortic root in a ratio of l:4. The solution was cooled to 8 to 12°C after cardiac arrest and infused for an additional five minutes. A cold, moderately hyperkalemic 3.5% glucose solution (solution B 11 in Table 3) and oxygenated blood in the same ratio were infused into the aortic root and venous grafts intermittently for 3 to 4 minutes at 15. to 20.minute intervals for maintenance of cardiac protection. Before declamping of the aorta the blood cardioplegia solution was warmed to 35 to 37°C and infused for 5 minutes. Before coming off CPB, an infusion of 5% glucose with NaCl40 mmol, K+ 76 mmol, phosphate 8 mmol, and MgC12 16 mmol per liter was started in both groups. This infusion continued at a rate of 1 mL/kg/h (glucose 0.05 gikgih) until the first postoperative morning. Additional potassium was given in IO- or 20.mmol increments if the serum K+ was lower than 4.0 or 35 mmol/L, respectively. Blood samples were taken and analyzed as indicated in the figures and tables. The MB fraction of creatine kinase (CK-MB)

Table 1. Data of Patients

Age (~4

R

55.9 (6.6)

54.9 (8.0)

1313

1412

Weight (kg)

78.6 (10.6)

76.7 (9.7)

Height (cm)

170.3 (10.5)

168.4 (6.3)

AMI 01112 (n) NYHA II/III/IV (n) EF LVEDP (mmHg)

Table 2. Preoperative Echocardiographic

Findings

GIK

R 10

Normal

7

Anterior hvpokinesla

3

I

Postenor hypokinesia

1

: 3

Inferior hypokinesia

3

Inferior akinesia

2

0

16

16

Total NOTE. Values are numbers of patients. Differences beween

groups were not significant (P = 0.607; ,yztest).

was analyzed postoperatively 2 to 2.5 hours after aortic declamping and on the first, second, and third postoperative mornings. Insulin concentrations were determined using a commercial kit (Diagnostic Products Corporation, Los Angeles, CA). The other laboratory tests were performed by routine methods. Systemic and pulmonary arterial pressures were recorded using a Kone M593 monitor (Kone Corporation, Instrument Division, Espoo, Finland). and data were collected as indicated in the tables. The cardiac output (CO) measurements were performed using a thermodilution catheter (7.5F-GSP 5107, Gould, Viggo-Spectramed, Oxnard, CA), and hemodynamic indices were calculated from standard formulae. Inotropic support was used to ensure a minimum cardiac index (CI) of 2.2 L/min/m* and a mean arterial pressure (MAP) greater than 60 mmHg. A 12-lead electrocardiogram (ECG) was taken preoperatively, immediately after arrival in the ICU, after 4 hours in the ICU, and on the following 7 postoperative mornings. Patients were on continuous ECG monitoring during the operation (leads 11, Vz) and during their stay in the ICU (lead II), and note was made of any heart rate. rhythm. or conduction disturbancea. and QRST changes. The statistical methods used were one-way analysis of variance for repeated measures, the nonpaired Student’s t test for nondependent variables. and the ,$ test for nonparametricvariables. Changes within the groups were analyzed using the paired Student’s I test with appropriate Bonferroni corrections. All data are given as means + SD. and the leve1 of significante was taken to be P < 0.05. RESULTS

The groups were comparable with regard to age, sex, weight, and preoperative cardiac status, as seen in Tables 1 and 2. Operative data are presented in Table 4. There was a tendency for a greater number of distal anastomoses in the GIK patients, and hence longer clamping times, but the differenccs were not statistically significant. NO differences were found between the groups regarding perfusion duraTable 3. Composition of Cardioplegic Solutions

GIK

Sex (M/F) (n)

KAUKORANTA, AND NUUTINEN

Solution B 1: KCI

50 mmol

Tribonat

30 mL

Ringer’s acetate

500 mL

Solution B ll:

51912

4/12/0

KCI

30 mmol

0114/2

ZIT311

NaCI

30 mmol

0.56 (0.05)

0.59 (0.07)

16.3 (5.7)

13.9 (4.9)

NOTE. Valuas are mean + (SD) or numbers of patients. All differences between groups were nonsignificant. Abbreviations: AMI, acute myocardial infarction in history; EF, left ventricular ejection fraction; LVEDP, left ventricular end-diastolic pressure; GIK, glucose-insulin-potassium

group; R, Ringer’s acetate group.

Tribonat

170 mL

CPD

250 mL

5% glucose

600 mL

NOTE. Tribonat (Pharmacia, Sweden) is a balanced buffer solution (trometamol

30 mmol, HCOB 16 mmol, phosphate

2 mmol, Na+ 21

mmol, acetate 20 mmol, aqua ster 100 mL, and pH 8.1). Abbreviation: CPD, citrate-phosphate-dextrose

solution.

‘)j -o- ‘>IIE) ‘-)(:as ‘1 ‘aral aas ‘At.’ pus 83 uaaMaaq sauy uaqorg :a&o~ .dnol6 u snsmA ~13 ~6.6 > d,,,, %qu~oru aywadowod )sq) ‘wd !n31 u! smoq p mye p3! !amaemdruaa (eaBuheqdoseu 3.6~ oa Buyuewal *e ‘~4 :uogelnuue3 3yoe je ‘e3 :e!saq~saue a~o~aq ‘!d :saru!) Buydrues poola ‘as T ueatu se uan! ave sanlel( ‘(-o-) H pue (-x_) ~10 sdnol6 Apnas u! suo!~er~ua~uo~ u!lnsu! runrag ‘2 6~

u,d

P3!

10” s! sdno,ô

uaawaq

pua 10’0 > .jr

awa,aga

:smo,,o,

(18’0 = d) lue3!)!U6!s ‘an,eh a,vle,adoa,d sns,a~ 90’0 > ,=!J

se paleo!pu!

a,e sdno,6

a,a san(e~ ‘y/loww

u! suo!ge,~ua~uo~ wn!sselod

(LZ’O) CiL’P

(81’0) 60’9

(ZYO) L6’E

(19’0) ÇL’P

~(96’0) PP’9

x(8L’O) 81’9

*(ZL’O) 19’9

*(86’0) 99’9

(09’0) PZ’P

(09’0) 80’9

UEE’O) 9L’E

x(LE’0) 89’E

(9E’O) 113

(6P’O) 903

(61’0) PL’P

(WO) 6L’P

(Pro) 913

(92’0) 913

P3!

131 dd MJ 1~. dq

1?3

od

!d

sazua,ag!a u,n,aS ‘g10N piod

t MI 48KM 4Prl31

sa6ueqD um!sseaod umas

W!

wqly

‘(OS) I ueaw se uay6

!d

>IIcJ

8

pa%awa saauaJag!p ~ueayy@s umoqs ave sa%ueqa ugnsu! umas

%!umma~ Su!.mp pue (-3/nur 9’0E1 T r.80~) uogelnuuea Bu!.mp sluaped air> aql u! suoyz~uaauoa q$y iasom@ pooIq aql se ulalled auws aql paMoqs sa8ueqa ugnsu! umag ~Llagwadoald ueql s[ar\al ~aqS!q le pau!ewa~ lnq ‘sdnol% qloq UI Lpel!tu!s [la3 suo!~e~~uaauoa asom pootq ‘ad3 lal3v .sdnol% qloq u! .wyu!s WM i?!~!uI.w~al 8upnp yead a!maa@hadAq aql ‘snqL ‘(7/1oum 1’~ T 0.~1) S?u! -LII.IBM~I 8uynp 8uymaao Iaaal lsaqS?!q aq1 qJ!M ‘ad3 8uyp dluo uaas SBM asuodsal ayaak+ad.iq ~ueay~ui?~s e ‘sluaged 8ynp (~/[oum z’p T 0.~1) [a~a[ yead 8 aql UI C?kqw.wma~ Mau e q]!M lnq ‘ad3 30 l.w]s aql Jaqe dolp .ma[a E kq paMo[ -103 alaM asaqL ‘(~/~ouuu p’s 7 p’zl) uoyelnuuea uodn paqaeal slaAal isaq%q aql ql!M ‘ad3 aIo3aq uo!sn3u! asoarqff %u!mp sluaged ~13 aql UI L~~ueay~ufd!sasol slaaal asom@ poolg wuaged air> aql u! slaAal Jaq%!q q]!M ‘uynsu! 103 10’0 > d pue aSOXq% -103g)‘o > d :sdno.G aq1 uaamlaq

au!welo,d

q 1 rl31 ,aqV

i%I!U,,aMa~ e!IUaqlOdAH ssedAq a~nu!~ 9 UO!lD”p”!-al,,

‘5 alqel

alaH ‘5. pur! 1 sBg u! pue asom@ poolq aqL

‘(01’0 = d) AIagaadsal

‘sluarled a pue >IIQ

“03 OOET fjE8 puB 622 i (jfjfj sI?M ‘asoan@ +I!WU03 ‘(11 a) uoyqos e@aldo!plea 30 lunowe IEJOJ aql pue yu 801 7 LEP SBM palals!u!wpt? uogn[os ~19 ssedkqald 30 awrqoA aqL .( s a[qeL) saiìwqa uysselod pue umpos umas UI 10 slaAal I!molewaq pue u!qo@ouraq poolq u! sdno_S aql uaarnlaq saaualag!p ou a1aM alaqL ~~ueay~u%!s JOU alaM saaualagrp asaql Inq ‘&.yduyaap Iaye do peq sluaged x x!s pw >II~ aizg pur! ‘el~oe aql30 Zu!dwep lage do 30 poyad IIoqs e peq sluaged a x!s pue swaped >IIE) mos ‘hzls n3I 30 uogmnp 10 ‘syaoqslalunoa 30 laqrunu ‘uogemp (AA) uogeII!~qy .ynayuah ‘uoy ‘pau!wo a,e sanleA

wosea,

lecyuq3al

,o,

slua!led asaql ql!m pal~a,,o~

paAelap

sem uo!le(l!,q!gap

t.j auo u! pue ‘po!,ad uo!lell!,q!),el”3!,lualzalnu!w-g WM

s!soluolseue

lels!p A,aua

xal)wn3,!3

:g,j3

$0 pua le papaau

,eln3!,luaA/uo!i3un~

,o, paisa,,o3 ,elnyluah

‘a,djA

!Awopa,avapua

1sodjA

‘1SOdjA

:alolsAse

3yoe

‘,,osj~

:alolsAse

‘adz

‘Vg !sasowoJsaue 40 s,aqlunu

pue

(P6) 9E

(P9) 6E

LI9IOL

Li916

1/1/&/11

0/0/9/1 L

(19’0) EP’Z

(68’0) LE’Z

iua!ied

‘suo!iell!,q!iap ,aye

auy

3!yoe

wo,,

!alu!g dluelz-sso,D

ge,6 lels!p ‘Vga

(OS) +

(EEL) E9

(SOL) 89

1

uo!lell!,q!)

‘UOlyd/ jA/dS

!dLuel>ap 3yoe a,o,aq

01 dwel3-sso,3

-sAse :ssedAq A,euowlndo!p,e3

IIV w,a!led

>llg auo UI,

~0 ,aqlunu

,ayeua3ed/dLue(3ap

3e!p,e3 )o A,aAoDa, snoaueluods

!(* aas) we,ave au!1 uo!lell!,q!, ,aln3!,lua~

ue Gu!,np palga,,o~

aql lua!ied

‘papaau suo!lell!,q!Jap ,dnol6 tj snsmt ~10 ~6.6 > ,j.* ‘dnolti ti snsmA ~19 ~6.6 > ,j* ~~U!UJOUIarr!gerado)sod asly ‘wd !n31 aqa u! slnoq 6 pue ‘p ‘1 mue ‘8_~3! !uo!aelas!u!tupe au!rueaord ral)s saanuy g, ‘dd !amaeraduw leaôuheqdoseu 3.6~ 0) 6uyu1e~a1 ae ‘m !e!uuaq)odAq asadaap )e ‘aq !ada JO saantqw 5 ral)8 ‘dq !uo!)elnuues -a!uoe le ‘es !AmBms a~o)aq ‘od !e!saqasaue alo~aq ‘!d :saurg Buydutes poola ‘QS T ueaw SB uarr!l ave san(eA ‘(-o-) u pue (-x-) ~10 sdnorô Apnas u! suo!)er)uamo3 asom poolg ‘L Bij

~lua3~yu6~suou a,am sdno,6 u,d

(P6) 8E

(99) 8E (86) 90

uealu

,aUe

uo!lel(!,q!) aw!l

‘al01

cyuoe ,330~

:suopa!,ta,qqV

uaannlaq sawa,aH!p a,e

sanlaA

Wp)

‘g10N

Am mi

L/Z/ t 10 swiwwa “oWd/jAldS *(aas) ,103jA (aas) isod jA (335) a,d jA (Das) alolsAsV

(P8) 18

(81’1) ÇL’Ç

(81’1) P6’P

LL/8

LI9

(LE) vol

(LZ) E6

(LP) L9L

(99) 99 L

9d3

(“!U’)

(U!W 33ov (,aqwnu/led) V3 wa

>IID

tl

eaea

angeradg

Atl3f)WlS

EZÇ

‘t a(qel

‘39V3

NI NOISfljNI

>llD SSVdABUd

524

WISTBACKA, KAUKORANTA, AND NUUTINEk

(166.8 ? 124.6 mUiL) with a postoperative return to preoperative levels. Thc R patients showed a significant hyperinsulinemic response only during rewarming (110.5 ? 59.9 mU/L). Free Fatty Acid and Ketone Body Chunges The changes al1 differed significantly (P < 0.01) between the groups (Table 6), with higher concentrations in the R patients. Serum FFA concentrations in the GIK patients were significantly lower during rewarming and on the first postoperative morning than before induction (P < 0.01); whereas those in the R patients were significantly elevated during cannulation (P < 0.05), but returned to the preoperative leve1 during rewarming. NO significant changes were seen in plasma acetoacetate levels in the GIK patients, but the R patients showed a tendency for a rise during cannulation and a significant elevation during rewarming (P < 0.01). Plasma hydroxybutyrate was below the detection limit in the GIK patients, but was elevated during cannulation and rewarming in the R patients. Serum CK-MB Changes CK-MB values (Table 7) were elevated postoperatively in both groups, but declined thereafter. Four GIK patients and three R patients had higher CK-MB values on the first postoperative morning than immediately after arrival in the ICU; but only two patients, one in each group, showed persistent ECG changes consistent with a myocardial infarction. Hemodynamic Changes The hemodynamic were no significant

responses are shown in Table 8. There differences between the groups with

Table 6. Serum Free Fatty Acid and Ketone Body Changes GIK

R

S-FFA Pre-induction

0.88 (0.54)

Cannulation

0.97 (0.31)

1.37 (0.57)x

Rewarming

0.37 (o.lo)t

0.81 (0.31)

ICU4

0.72 (0.26)

0.79 (0.24)

ICUpm

0.34 (o.lo)t

0.55 (0.25)* 0.13 (0.08)

0.90 (0.51)

P-Acetoacetate Pre-induction

0.10 (0.05)

Cannulation

0.09 (0.07)

0.18 (0.15)

Rewarming

0.16 (0.12)

0.31 (0.18)t

ICUl

0.08 (0.11)

0.18 (0.18)

ICUpm

0.11 (0.09)

0.13 (0.12)

P-Hydroxybutyrate Pre-induction

0.00 (0.00)

0.01 (0.03)

Cannulation

0.00 (0.00)

0.16 (0.14)t

Rewarming

0.00 (0.00)

0.32 (0.23)t

ICUl

0.00 (0.00)

0.01 (0.03)

ICUpm

0.01 (0.03)

0.00 (0.00)

NOTE. Serum FFA, plasma acetoacetate,

and hydroxybutyrate

con-

centrations in mmol/L before induction, at aortic cannulation, during rewarming phase of CPB, at 1 or 4 hours after arrival in ICU, and on the first postoperative morning. Values are given as mean 2 (SD). Differences within groups are indicated as fellows: *P < 0.05 and tP < 0.01 versus preoperative value.

Table 7. CK-MB Fraction Changes GIK

R

Postop

4% (18)

49 (20)

Day 1

33 (37)

33 (44)

Day 2

51111

8 (16)

Day 3

2 17!

3 17)

NOTE. Values are mean k (SD) and unitsil. NO differences were seen between the groups: P = 0.788. Abbreviations: Postop, immediately after arrival in ICU; Day 1, 2 and 3, first, second, and third postoperative mornings, respectively.

regard to heart rate, mean systemic and pulmonary artery pressures, filling pressures, CI, or systemic and pulmonary resistances. Inotropic Support A total of 11 GIK patients and 12 R patients needed dopamine support (P = 0.95; x2 test) (Table 9), and two in each group were also given smal1 doses of norepinephrine (17 to 107 ng/kg/min). Another patient in group GIK received only a norepinephrine infusion. Two patients, one in each group, needed only a dobutamine infusion, the doses being 3.2 to 7.9 pg/kg/min. A total of 13 patients in each group needed inotropic support at some stage during the follow-up period, but none needed intra-aortic balloon pump support. Electrocardiographic Changes Only one GIK patient and one R patient needed a temporary pacemaker at the end of bypass. The GIK patient developed a stable sinus rhythm within 30 minutes after CPB and the R patient on arrival in the ICU. One further GIK patient and four R patients got temporary pacemakers during reperfusion because of an initial atrioventricular black, bradycardia, or nodal rhythm, but these patients were in sinus rhythm by the end of CPB. In each group, 25% of the patients (4 of 16) had at least one episode of atrial fibrillation (AF) during the first postoperative week, but only one R patient with preoperative AF had persistent AF. Partial bundle-branch blocks were seen in hvo GIK and two R patients after arrival in the ICU, but these changes had disappeared on the first postoperative morning; one other R patient had a residual left bundle-branch black on the seventh postoperative day. These changes all normalized in a repeat ECG 1 month later. Ventricular premature beats in connection with moderate hypokalemia were seen in the ICU in two GIK patients and one R patient; and in one GIK patient (K+ 3.4 mmol/L), the premature beats were associated with a short period of ventricular tachycardia. Therapy consisted of potassium correction and lidocaine infusion. One patient in each group had permanent QRST changes and CK-MB enzyme elevations consistent with a perioperative myocardial infarction. Both infarctions were uncomplicated and the patients made uneventful recoveries without hemodynamic problems. A further two GIK patients and one R patient with slight CK-MB enzyme elevations on the first postoperative morning manifested transient ST seg-

525

PREBYPASS GIK INFUSION IN CABG SURGERY

Table 8. Hemodynamic Parameters ICU4

ICUpm

88 (14)*

90 (12)X

88 (12)*

85 (9)*

95 (16)’

92 (16)*

74 (6)*

82 (1l)t

66 (7)*

75 (9)’

73 (10)’

91 (19)

75 (15)X

74 (7)s

6.7 (2.2)

9.5 (3.0)X

8.6 (2.4)t

8.2 (1.6)t

8.6 (2.7)

8.5 (2.6)

9.9 (2.5)

8.4 (2.8)

8.7 (2.4)

8.4 (2.9)

16.8 (3.1)

19.8 (3.5)t

20.1 (3.9)t

18.8 (3.2)

18.2 (3.6)

19.8 (3.6)

20.2 (3.1)

19.6 (4.3)

19.4 (4.4)

17.6 (4.2)

10.9 (2.1)

ll.3

12.1 (3.7)

12.0 (1.9)

3.2 (0.6)

3.2 (0.6)

3.3 (0.9)

3.1 (0.7)

1165 (275)

883 (228)*

1265 (294)

1242 (333)

918 (255)*

1460 (466)

HR (beats/min) GIK R MAP (mmHg) GIK R CVP (mmHg) GIK R MPAP (mmHg) GIK R PCWP (mmHg) GIK R Cl (L/min/m2) GIK R SVR (dyne . GIK

s

s

Postbp

61 (11)

88 (12)*

61 (10)

87 (15)X

93 (12) 98 (12)

(3.0)

ICUl

9.0 (2.6)t 10.1 (2.5)

8.7 (2.1)*

9.4 (2.3)

10.2 (1.8)

9.4 (2.3)t

2.6 (0.6)t

2.8 (0.5)

3.0 (0.5)

2.5 (0.6)t

2.7 (0.4)t

3.1 (0.8)

932 (327)t

965 (330)t

cmm5)

R PVR (dyne

Preop

1123 (356)

972 (282)t

cmm5)

GIK R

80 (41)

115 (77)

195 (85)*

156 (50)’

126 (48)*

106 (44)

118 (39)

162 (74)t

149 (54)t

118 (45)

NOTE. Hemodynamics before induction, after CPB and at ICU at 1 and 4 hours after arrival and on the first postoperative morning. Values are given as mean i (SD). Abbreviations:

HR. heart rate; MAP, mean arterial pressure; CVP, central venous pressure; MPAP, mean pulmonary artery pressure; PCWP,

pulmonary capillary wedge pressure; Cl, cardiac index; SVR, systemic vascular resistance; PVR, pulmonaryvascular Differences within groups are indicated as fellows: *P

< 0.01

and tP

resistance.

< 0.05 versus preoperative value. Differences between groups were not

significant.

ment changes depressions.

without

any persistent

Q waves or R wave

DISCUSSION

A number of investigations have been carried out over the last two decades into strategies for supportive cardiac nutrition aimed at increasing cardiac glycogen, ATP, and creatine phosphate reserves and enhancing the utilization of these by triggering the glycolytic enzyme systems before the ischemic event (eg, aortic cross-clamping).12J3~19-25 Infusion of a GIK solution has been used for this purpose. However, much of the work has involved patients undergoing valvular operations,12J3*21-23 using crystalloid cardioplegia protection during aortic cross-clamping of relatively limited duration. There have also been great technical Table 9. Dopamine Support (pglkglmin) GIK

R

N

Dose

N

DOE

Postbp

5

4.7 (0.9)

7

3.7 (1.3)

ICUI

7

4.0 (2.2)

7

3.5 (0.8)

ICU4

7

4.7 (2.1)

10

5.1 (2.0)

ICU8

6

6.0 (1.9)

11

4.8 (1.7)

ICUpm

9

4.3 (2.2)

10

3.8 (1.5)

NOTE. Dopamine support (pg/kg/min)

after CPB and in ICU at the 1,

4, and 8 hours after arrival and on the first postoperative

morning.

Values as number of patients and doses are given as mean ? (SD).

variations with respect to glucoselinsulin ratios and infusion rates. The infusion of GIK requires frequent laboratory measurements of blood glucose and electrolyte status in order to avoid hyperglycemic and/or hypoglycemie episodes with ensuing serum electrolyte and fluid balance disturbances.26327The most important argument for avoiding hyperglycemia in cardiac surgery is its possibly detrimental effects on neurologie outcome in connection with cerebral ischemia.15,** For several years, a GIK solution was routinely used for both valve reconstruction and CABG patients in this institution. The glucose load before CPB (0.6 g/kg/h) corresponded to that mentioned in some earlier reports.20-23 There is a wide variation in glucose/insulin ratios in the previous reports, and the insulin infusion rate used here was relatively smal1 (0.12 U/kg/h). The higher blood glucose and serum insulin levels in the GIK patients before CPB were expected, but there was great interindividual variation, with the maximum blood glucose concentration of 20.1 mmol/L measured upon cannulation (male, 55 yr), and the two lowest concentrations, 2.3 and 2.0 mmol/L, found after starting CPB (both males, 61 and 63 yr). These low levels are probably partly due to hemodilution, but also to an insulin effect, because both these patients had lower glucose concentrations during cannulation than at the start of surgery. Thus, the amount of insulin given was too smal1 to attenuate the

526

WISTBACKA,

hyperglycemic response to glucose infusion and surgical stress in most patients, but was high enough to reduce glucose concentration to a potentially dangerous leve1 in some individual cases. Hemodilutional hypoglycemia of short duration after starting CPB has been reported previously when a glucose-free priming solution was used,?” but this was not seen in the present R patients whose blood glucose levels ranged between 4.2 and 7.3 mmol/L after starting bypass. Thus, GIK infusion was associated with a considerable risk (2116 = 12.5%) of hypoglycemia, which may have detrimental effects on the central nerveus system. There were no differences between the GIK and R patients in the hyperglycemic peak response during rewarming, and blood glucose concentrations rose to a potentially harmful leve1 in both groups. Hyperglycemia during CPB was mainly due to a genera1 surgical stress response,3’t but probably was also secondary to the glucose infusion given with the cardioplegia in both groups. The prebypass insulin given to the GIK patients was insufficient to attenuate these effects during hypothermia, when endogenous insulin secretion is depressed.3’.32 The elevated insulin levels observed in both groups during rewarming are mainly dut to the hyperglycemia-induced stimulation of endogenous insulin secretion? and, to a lesser extent, to exogcnous insulin in the GIKpatients (TL/? of insulin only 7 to 15 minutes). Although the amount of insulin given was relatively small, it was effective enough to give significantly elcvated serum insulin concentrations in the GIK patients. A clear depressant effect was also seen on lipolysis and ketogenesis, with signifìcantly lower serum FFA and plasma acetoacetate and hydroxybutyrate concentrations during cannulation and CPB. There were, nevertheless, no clinically detectable cffects on VF frequency and duration during cross-clamping of the aorta and after declamping, or on the frequency of postoperative arrhythmias. The frequency of arrhythmia was relatively low in both groups and comparable to rates reportcd elsewhere.lJ This may be attributed to good myocardial protection during aortic cross-clampand possibly to the postoperative supplementaing, 18~34~3s tion of glucose, potassium, and magnesium in both groups.“h,X7 The number of anastomoses was higher and the duration of cross-clamping and perfusion longer in these cases than has been described in earlier assessments of GIK involving

KAUKORANTA.

AND NUUTINERI

CABG paticnts.‘“,Z’ :’ Althouph it haa been suggestcd 111;1i diffcrences in clinical appearance bctwccn trcatcd and nontreated groups are more likcly to occur with longcr cross-clamping times,:’ no such dilfcrcnccs could bc shown hcrc with rcgard to myocardial injury either immediatelh after CPB or in thc postopcrativc period. The CK-MB enzymc fractions were elevatcd to a similar degree in both groups, and the frequency of postopcrative ECG change. was similar. Also, the incidcncc of perioperativc myocardial infarction was 6.25p& in both groups. Thus. the differcncc between groups was O%>,although with a 95% confidcncc interval of -16.8 to + 16.8%. refecting the smal1 size of thc groups. Hence, conclusions should be drawn with caution. The GIK infusion time before CPB was relatively short in this trial. Previous infusion times have been longer,‘2.~“.“,~“.?i the insulin doses grcater.‘3-‘l-‘~,iK.‘” and/or infusion bas been given during and after ischcmia. Consequently. direct comparison hctween the present fìndings and previous results is difficult. The total numbcr of patients given inotropic support was high, 13 patients in each group, a fact that can be attributed to the relatively high degree of atherosclerosis. high cndartercctomy frequency, large number of distal anastomoses. and long aortic cross-clamping times. Mean doses were small, however, and the inotropic support could be stopped on the following day in all patients. Coming off CPB was uneventful in ah cases, and no patient needed intra-aortic balloon counterpulsation. It rcmains to be shown whether an overnight GIK infusion before surgery combined with perioperative and postoperativc GIK infusion ratios and rates that come closer to the optimum could be of benefit to CABC patients when blood cardioplegia is used for cardiac protection. Thc present trial suggests that the critical moments with respect to glucose concentration changes are the commencement of CPB. associated with a combincd risk of hemodilutional hypoglycemia and insulin-induced hypoglycemia, and rewarming associated with a major hyperglycemic stress response. Therefore. any attempt at metabolic manipulation should be undertaken cautiously. GIK therapy necessitates careful monitoring of blood glucose and electrolyte concentration changes throughout the operation. Finally, glucose infusion should not be stopped abruptly, in order to avoid rebound insulin-induced hypoglycemia.

REFERENCES 1. Nikki P, Takki S, Tammisto T, Jäáttelä A: Effect of operative stress on plasma catecholamine levels. Ann Clin Res 4:146-351, 1972 2. Walsh ES, Paterson JL, O’Riordan JBA, Hall GM: Effect of high-dose fentanyl anaesthesia on the metabolic and endocrine response to cardiac surgery. Br J Anaesth 53:1155-1165, 1981 3. Clarke RSJ: The hyperglycaemie response to different types of surgery and anaesthesia. Br J Anaesth 42:45-52, 1970 4. Äärimaa M, Syvälahti E, Viikari J. Ovaska J: Insulin, growth hormone, and catecholamines as regulators of energy metabolism in the course of surgery. Acts Chir Stand 144:411-422,197s 5. Stjernström H, Jorfeldt L, Wiklund L: The influence of abdominal surgical trauma upon the turnover of some blood-borne energy metabolites in the human leg. J Parenteral Enteral Nutr 5:207-214. 1981

6. Ekroth R. Nilsson F. Berggren H. et al: Insulin sensitivity and glucose uptake in the course of surgical treatment for valvular aortic stenosis. Stand J Thor Cardiovasc Surg 16:137-140, 1982 7. Askanazi J. Van de Wiele B, Rosenbaum SH: Nutrition and cardiac function. in Kaplan J (ed): Cardiac Anesthesia (ed 2). vol 2. Orlando, FL, Grune & Stratton, 1987, pp 1077-1103 8. Teoh KH, Mickle DA. Weisel RD, et al: Decreased postoperative myocardial fatty acid oxidation. J Surg Res 44:36-44. 1988 9. Henderson AH, Craig RJ, Gorlin R, Sonnenblick EH: Free fatty acids and myocardial function in perfused rat hearts. Cardiovast Res 4:466-472, 1970 10. Liedtke AJ, Nellis S, Neely JR: Effects of excess free fatty acids on mechanica1 and metabolic function in normal and ischemie myocardium in swine. Circ Res 43:652-661, 1978 I 1. Tansey MJ, Opie LH: Relation between plasma free fatty

PREBYPASS

acids and myocardial

GIK INFUSION

527

IN CABG SURGERY

arrhythmias within the first twelve infarction. Lancet 2:419-422,1983

hours

of acute

12. Oldfield GS, Commerford PJ, Opie LH: Effects of preoperative glucose-insulin-potassium on myocardial glycogen levels and on complications of mitral valve replacement. J Thorac Cardiovasc Surg 91874-8781986 13. von Bormann B, Scheld HH, Podzuweit T, et al: Enhancement of myocardial energy potentials in man by glucose-insulin treatment before and after ischaemic heart arrest. J Cardiovasc Surg 26:182-186,198s 14. Coffey CE, Massey EW, Roberts KB, et al: Natura1 history of cerebral complications of coronary artery bypass surgery. Neurology 33:1416-1421, 1983 15. Myers RE, Yamaguchi S: Nervous system effects of cardiac arrest in monkeys. Arch Neurol34:65-74, 1977 16. Lanier WL, Stangland KJ, Scheithauer BW, et al: The effects of dextrose infusion and head position on neurologie outcome after complete cerebral ischemia in primates: Examination of a model. Anesthesiology 66:39-48, 1987 17. Sieber FE, Smith DS, Traystman RJ, Wollman H: Glucose: A reevaluation of its intraoperative use. Anesthesiology 67:72-81, 1987 18. Buckberg GD: Strategies and logie of cardioplegic delivery to prevent, avoid, and reverse ischemic and reperfusion damage. J Thorac Cardiovasc Surg 93:127-139,1987 19. Kappelman MD, Hewitt RL: Protection of the ischemic heart with energy substrate and potassium during cardiopulmonary bypass. Surg Forum 25:153-155,1974 20. Salerno TA, Wasan SM, Charrette EJP: Glucose substrate in myocardial protection. J Thorac Cardiovasc Surg 79:59-62, 1980 21. Berggren H, Ekroth R, Herlitz J: Improved myocardial protection during cold cardioplegia by means of increased myocardial glycogen stores. Thorac Cardiovasc Surg 30:309-329,1982 22. Haider W, Schütz W, Eckersberger F, et al: Optimierung des myokardialen Energiopotentials durch präoperative hohe Insulinzufuhr als Myokardprotektion in der offenen Herzchirurgie. Anaesthesist 31:377-382, 1982 23. Haider W, Bentzer H, Schiitz W, et al: Improvement of cardiac preservation by preoperative high insulin supply. J Thorac Cardiovasc Surg 88294-300, 1984 24. Lolley DM, Myers WO, Ray JF, et al: Clinical experience with preoperative myocardial nutrition management. J Cardiovasc Surg 26:236-243,198s 25. Berggren H, Ekroth R, Hjalmarson A, et al: Enhanced myocardial protection from preoperative carbohydrate infusion in

addition to maintained beta-blockade. J Cardiovasc Surg 26:454456,1985 26. Bohrer H, Fleischer F, Krier C: Hyperkalemic cardiac arrest after cardiac surgery following high-dose glucose-insulin-potassium infusion for inotropic support. Anesthesiology 69:949-953, 1988 27. Marwick TH, Woodhouse SP: Severe hypophosphataemia induced by glucose-insulin-potassium therapy. A case report and proposal for altered protocol. Int J Cardiol3:327-330,1988 28. Lanier WL: Glucose management during cardiopulmonary bypass: Cardiovascular and neurologie implications. Anesth Analg 72:423-427,199l 29. McKnight CK, Elliott MJ, Pearson DT, et al: Continuous monitoring of blood glucose concentration during open heart surgery. Br J Anaesth 57:595-601,1985 30. Stanley TH, Berman L, Green 0, et al: Plasma catecholamine and cortisol responses to fentanylioxygen anesthesia for coronary artery operations. Anesthesiology 53:250-253, 1980 31. Moffit EA, Rosevear JW, Molnar GD, McGoon DC: Myocardia1 metabolism in open heart surgery. Correlation with insulin response. J Thorac Cardiovasc Surg 59:691-706,197O 32. Allison SP: Changes in insulin secretion during open heart surgery. Br J Anaesth 43:138-143,197l 33. Kuntschen FR, Galletti PM, Hahn C: Glucose-insulin interactions during cardiopulmonary bypass. Hypothermia versus normothermia. J Thorac Cardiovasc Surg 91:451-459, 1986 34. Rosenkranz ER, Vinten-Johansen J, Buckberg GD, et al: Benefits of normothermic induction of blood cardioplegia in energy-depleted hearts, with maintenance of arrest by multidose cold blood cardioplegic infusions. J Thorac Cardiovasc Surg 841667-676, 1982 35. Teoh KH, Christakis GT, Weisel RD, et al: Accelerated myocardial metabolic recovery with terminal warm blood cardioplegia. J Thorac Cardiovasc Surg 91:888-895,1986 36. Aglio LS, Stanford GG, Maddi R, et al: Hypomagnesemia is common following cardiac surgery. J Cardiothorac Vast Anesth 5:201-208, 1991 37. Finlayson DC: Magnesium: Its time has come. J Cardiothorac Vast Anesth 5:199-200, 1991 (editorial) 38. Svedjeholm R, Hallhagen S, Ekroth R, et al: Dopamine and high-dose insulin infusion (glucose-insulin-potassium) after a cardiac operation: Effects on myocardial metabolism. Ann Thorac Surg 51:262-270,199l 39. Gradinac S, Coleman GM, Taegtmeyer H, et al: Improved cardiac function with glucose-insulin-potassium after aortocoronary bypass grafting. Ann Thorac Surg 48:484-489, 1989