dcla Anaeslhesiol Scand 1991 : 35: 2 16-220

Myocardial effects of adenosine- and sodium nitroprusside-induced hypotension: a comparative study in patients anaesthetized for abdominal aortic aneurysm surgery A.

OWALL and A. SOLLEVI

Department of Cardiothoracic Anaesthetics and Intensive Care, Karolinska Hospital and Department of Anaesthesiology, Sljdersjukliusct. Stockholm, Sweden

The effects of adenosine and sodium-nitroprusside (SNP) on central and myocardial haernodynamics and metabolism were evaluated during fentanyl anaesthesia (100 p g ' kg-') in six patients with peripheral vascular disease. The investigation was performed during stable anaeshtesia, before scheduled abdominal aortic gralt surgery. Adenosine and SNP were infused intravenously in random order over 20 min, leaving a 30-rni11 control period in between. The vasodilators were titratcd in order to reduce mriin ertcrial pressurc I,v approximately 250/,,. Adenosine (9Ok 20 pg. k g - ' . min-I) reduced mean arterial pressure from 10.9 f 0.3 t o 8.4k0.4 kPa ( 8 2 f 3 t o 6 3 k 3 mrnHg), a n d S N P (0.7_+0.1pg.kg-'.mirir') from tl.Of0.2 to8.4fO.U k k i (83 & 3 mmHg to 63 3 mmHg) during the hypotension period. Cardiac index remained unaffected during induced hypotension with both vasodilators, while heart rate increased during SNP infusion (8 i U",,j ; t i i d remained unafFected with adenosine. Left ventricular stroke work index and myocardial oxygen consumption decreased during SNP infusion (33 3"j0 and 17 & 50i0, respectively), while these parameters were unchangcd with adenosine. Adenosine hypotension increased coronary sinus flow 1-2 fold (128 k 26";,),togt*thcr with increased coronary sinus oxygen content (96 k 1 IYL). I n contrast, coronary sinus flow decreased during SNI' hypotension ( - 15 f 40/,) with unaffected coronary sinus oxygen content. It is concluded that adenosine, i n contrast to SNP, is associated with a hyperkinetic myocardial circulation. Received 20 March, acctpted for publication 22 August 1990

Kpr words: Adenosine; coronary circulation; hypotension, controlled; nitroprusside

Controlled hypotension may be employed to reduce intraoperative haemorrhage, and the vasodilators adenosine and sodium nitroprusside (SNP) have been used clinically for this purpose (1, 2). Many organs including the heart have a vascular autoregulatory mechanism that maintains sufficient blood supply in a wide range of perfusion pressures ( 3 ) . Below a certain blood pressure level, organ perfusion may be altered and pressure dependent. This critical level of perfusion pressure varies between organs and may be different among individuals. The autoregulatory mechanisms may also be affected by agents acting as powerful vasodilators (4). Adenosine has recently been introduced as an agent to induce controlled hypotension in humans (1, 5). Induced hypotension (mean arterial pressure 5.3-6.7 kPa, 40-50 mmHg) is characterized by reduced systemic vascular resistance in parallel with increased cardiac output while the heart rate response is minor ( 1, 5). Adenosine-induced hypotension is remarkably stable and rebound hypertension is not seen ( 5 ) . This

stable effect may be explained by the unaffected levels of circulating catecholamines during anaesthesia a s well as the unchanged plasma renin activity that has been demonstrated during adenosine-induced hypotension (6). The myocardial effects ol' adenosine infusion in man are characterized by marked coronary vasodilation with increased coronary sinus blood llow without signs of increased myocardial work (6, 7). SNP induces arteriolar vasodilation and venodilation, and SNP is widely used for controlled hypotension and afterload reduction (8). The cardiovascular effects of SNP in anaesthetized patients are characterized by reduced systemic vascular resistance. Both maintained and reduced cardiac output and cardiac filling pressures have been reported (2, 9). Coronary sinus blood flow is not affected during SNP administr'ition to awake patients (10). Possible disadvantagrs are tachyphylaxis, rebound hypertension and risk 0 1 cyanide poisoning ( 1 1-1 4). The purpose of this study was to compare the central and myocardial haemodynamic etlects of hypotension

MYOCARDIAL EFFECTS O F ADENOSINE AND NITROPRUSSIDE

indiic.ed by adenosine and SNP. The drugs were compared in a group of anaesthetized patients, with peripheral vascular disease, at a dose sufficient to reduce mean arterial pressure to approximately 8 kPa (60 mmHg) . The comparison was therefore undertaken in random order in the same anaesthetized patient, thereby evaluating myocardial circulatory and metabolic effects during hypotension induced by the two vasodilators adenosine and SNP. Data concerning the myocardial effects of adenosine in these subjects have previously been partially reported (7). PATIENTS AND METHODS Six piitients scheduled for elective abdominal aortic aneurysm surgery wereinvcstigated. The study included patients with hypertension atrial fibrillation, Demographic and clinical data are given in Table 1. Preoperiltiw exercise testing was not performed in any case. Patients with angina prctoris or a previous myocardial infurtion were not included. The study was undertaken during stable anaesthesia prior to surgery. All nirdication was discontinued 8 h before the study began. 'I'ht. protocol was approved by the Ethical Committee of the Karolinska Hospital and patients were included after obtaining their written inforrried consent. Intramuscular premedication with morphine 0. I5 mg kg and scopolamine 6 pg . kg- was given 30 min prior to anaesthesia. Anaesthesia was induced and maintained with fentanyl 100 pg . kg ' _Patients wereintubated after pancuronium 0.1 mg. kg-I and they were mechanically ventilated (Engstrom 2000, Sweden) with a mixture of oxygen in air, and Fio, was continuously monitored (0.39i 0.01, s.e.mean) (Eliza, Engstrom, Sweden). Ringer's solution (2 mt . k g - ' . h-I) was given as preoperative fluid. No additional drugs or fluids were given throughout the investigation. Following induction of anaesthesia a catheter was inserted into the left rAdial artery ( 1 .O mm Teflon cannula), and a flow-directed, quadruplr-lumen pulmonary artery catheter (Edwards, model 93 A-8317.51.: \'I P) was inserted via the left subclavian vein. A thermodilution cathrtvr j 7.5F, Webster Laboratories, Altadena, California) was insertrd into the coronary sinus via the right internal jugular vein, under fluoroscopic control and continuous pressure recording. Correct posi-

'

\r.,irs

Sex Drugs

tion in the coronary sinus was verified by analysis of blood oxygrn saturation. Arterial, pulmonary artery, and right atrial pressures wcrt. recorded continuously by transducers placed at the midthoracic level. Mean pulmonary capillary wedge pressure was recorded intermittently. Cardiac output was measured by triplicate determinations, using 10 ml of cold saline ( 2 1°C) injected at end-expiration. Coronary sinus blood flow was determined in triplicate by the retrogradr thermodilution technique, using intermittent infusions of saline ill room temperature during 20s at a rate of 33 ml . min-I. The flows reported represent a mean of the triplicate determinations. A standard V,-lead ECG was intermittently recorded on a Mingograph" (Siemens, W Germany) recorder (50 m m ' s - ' ) in all patients. Arterial, mixed venous, and coronary sinus blood samples were obtained simultaneously for blood gas and oxygen content analysis. Oxygen and carbon dioxide tensions, and p H were measured with an ABL 3 blood gas analyzer (Radiometer, Denmark). Haemoglobi~i oxygen saturation was obtained using an OSM 3 Hemoximeter' (Radiometer). Haemoglobin concentration was also determined i l l each sample. Arterial and coronary sinus blood lactate levels were measured according to Tfelt-Hansen & Siggaard-Andersen ( 15). Patients were allowed 20 min rest after catheterization. Adenosine (5.3 m g . ml-' clinical solution from the pharmacy ol the Karolinska Hospital) and SNP (Niprid@, Hoffman-La Rochc.. Basel) were administered in random order as a continuous infusion into the superior vena cava. Measurements and blood sampling werr performed prior to the infusion of adenosine and SNP and at 20 miii ofstable arterial blood pressure reduction. There was a 3 0 4 1 1 rest in between the administration of the two vasodilators. Calculations and statistics Systemic vascular resistance index was calculated as (mean arterial pressure - mean right atrial pressure) / cardiac index. Left ventricular stroke work index was calculated as (mean left ventricular systolic pressure mean pulmonary capillary wedge pressure) . stroke volumr index, where mean left ventricular systolic pressure was calculated from the arterial pressures (16). Coronary vascular resistance was estimated as (diastolic arterial pressure - mean pulmonary capillar? wedge pressure) / coronary sinus blood flow. Results are expressed as mean standard error of the mean. Data were analysed using the Wilcoxon signed rank test for paired data, comparing each control with the effect of the drug. Significant probability values were accepted at Plr2

Sysrrmic and rnyorardinl harmodynamic variables at basal conditions and during adenosine- and sodium nitroprusside-(SNP)-induced hypottmsioii i n six frtttanyl-anarstlirtizrd patients. M ~ a n f s s.e.mran. * = P < 0.05, significantly different from basal.

Mran artrrial blood pressure kPa jmmHg) 1)iastolic arterial blood pressure kl'a j nimHg I i l r a n pulmonary artrry prc-ssurr kPa [ mrnHgi h,lr;in pulmonary capillary wrdgr pressurr kPa irnmHq) hlran right atrial prt"ssurr kPa ( mmHq) Hr;irr ratc brats . min (hrdiac indrx I min I . m Srrokr volumr indcx ml n - ' Sysrcmic vascular rrsistancr indcx kPa I - ' . min. 111' (:orimary sinus I h o d flow ml miti-' (:oronnry vascular rcsistance kl'a . I I min Lcli vrntricular strokr work index.] . m - 2

'

~'

nary sinus blood flow. Adenosine caused an increase i n flow of 128+ 26OO while SNP reduced coronary sinus flow by 15 f 4O, (Fig. 1 and Table 2). Concomitantly, adcnosine reduced calculated coronary vascular resistance (64 + 3"/,, Table 2) and increased coronary sinus oxygen content (96 f 1 l o ~ oTable , 3 ) . Myocardial oxygen consumption was thereby maintained with adenosine but reduced with SNP (Table 3 ) . Arterial and coronary sinus lactate concentrations were not significantly affected by induced hypotension ('Table 3 ) . Myocardial net lactate production did not occur on any occasion. However, coronary sinus lactate concentration increased in live patients and remained unchanged in one patient during adenosine infusion while coronary sinus lactate concentration increased in three patients and decreased in three patients during SNP infusion. The ECG revealed no ST-T segment de-

Basal

During adenosine

10.9f0.3 (82 f 3) 8.1f0.4 (61 f 3) 2.3f0.2 (17f2) I .6 f 0.2 ( 1 2 ? 1) 0.9 f 0. I (7f 1) 63 f 4 2.1fO.2 33f 1 5.0f0.3 107f6 62 f 6 0.41 f 0.01

8.4f0.4* (63 f 3.) 6.4f0.4* (48f 3') 2.2fO.2 (16f 1) I .4 0.2 (11 f I ) 1.OfO.1 (8f I ) 65 f 3 2.5f0.3 37 f 3 3.2f0.3* 236f 18, 22f 1' 0.37 f0.03

Basal

11.OfO.2 (83 f 2) 8.2k0.4 (61 f 3) 2.2f0.2 (16f2) I .6 f 0.2 (12f2) 0.8f0.1 ( 6 f 1) 66f3

*

2.1iO.2 32f2 5.0f0.4 108+ I I 64f 7 0.41 f 0.02

During SNP

8.4+0.3* (63 f 3*) 6.7?0.4* (50 f 3*) 1.9+0.2* ( l 4 f 2') I .2 f 0.2* (9 f 2*) 0.6fO.l ( 5 f 1) 71f3* 2.1f0.2 29 f 2* 4.0f0.4' 89 f 6* 63 f 6 0.27 f 0.02*

pression of more than 2 mm in any patient at basal conditions or during hypotension (data not shown). DISCUSSION Both adenosine and SNP were equally effective in reducing arterial pressure by 25% in these patients. Cardiac output remained unaffected with both agents and hypotension was caused by peripheral vasodilation. Myocardial filling pressures remained unaffected during adenosine infusion, while SNP reduced the pulmonary capillary wedge pressure. This may reflect differences in the site of vasodilatory action within the arterial and venous vascular beds. Hypotension is a stimulus that activates the baroreceptors. Thus, SNP-induced hypotension is frequently accompanied by reflex tachycardia ( 14, 17), and in-

'l'al)lr 3 Systrmic and myticardial mrtabolic vsriablrs at basal ronditiona and during adenosine- and sodium nitroprusside-( SNPiinduced hypotension i n six li-ntanyl anarsthrrized patients. Mr;ins f s.r.mean. * = P i 0.05, significantly different from basal. ~

Basal Artcrial 1'0, kl'a Artcarid Poo, kPa .4rtrrial oxygen content ml 1 Coronary sinus oxygen cunteilt nil . I I Artrrio-pulmonary artpry oxygrn cnnrent differenrr ml 1Artrrio-coronary sinus oxygeii rontcnt dilli.rrncr nil . I - ' \\'hole body oxygrn ronsumption ml . min I Myocardial oxygen consumption ml . min Artrrial Iartarc conrentration nimol . I - ' (iironary sinus lactatc ronrentration mmol I - I ~

'

~

~

~

'

I

14.7f 1 . 1 4.5 f 0.3 157 f 9 52 f 4 48 f 2 106f6 179f 13 11f1 0.7 f 0. I 0.5 f 0.02

~~~

During adenosine 1 I .9 f 0.8

4.6 f 0.3 152 f 8* 101 f 8* 38 f 3* 52 ?: 4* 168f 19 12f I 0.7 f 0. I 0.6 0. I

*

~

Basal

During SNP

14.22 1.0 4.6 f 0.3 I57 f 8 55f4 47 f 3 103f7 1 7 7 f 10 IIf I 0.7 f 0. I 0.5 f 0.04

12.0f0.7 4.7 f 0.3 151 + 9 * 52 f 4 47 f 3 99 6 171 f I I 9 f I* 0.7 f 0.1 0.5 f 0.1

219

MYOCARDIAL EFFECTS O F ADENOSINE AND NITROPRUSSIDE

ml/min -I

250

-

200

-

/ / / /

150

/

-

/ / /

100

50 0

-

1

, Before hypotension

0 Adenosine

,A

SNP

,

During hypotension

Fig. I . Coronary sinus blood flow (rnl/min) before and during adenosine- and sodium nitroprusside-(SNP) induced hypotension in six fentanyl-anaesthetized patients.

creased heart rate was observed during SNP infusion. In contrast, heart rate was not significantly increased during the same level of adenosine hypotension. The chronotropic response to adenosine in man is complex. In conscious man increased heart rate has been reported during adenosine infusion at a similar dose range (18, 19). This dose-dependent positive chronotropic response may be caused by a combination of vagal withdrawal and sympathetic stimulation (18). A dose-dependent and transient negative chronotropic effect of adenosine is seen in man following bolus injection (20, 21), since higher concentrations of adenosine exert a direct inhibitory effect on the pace-maker cells of the sinus node (22, 23). Thus, a direct negative chronotropic effect of adenosine did probably counteract a positive chronotropic response that was mediated via hypotension-induced baroreceptor stimulation. Marked differences in myocardial haernod ynarnics were observed between the two vasodilators. The rapid and large increase in coronary sinus blood flow during adenosine infusion is consistent with many reports that adenosine is a potent coronary vasodilator (24, 25). A preferential coronary vasodilatory effect of adenosine compared to other vascular beds has previously been described in dogs using microspheres and in patients undergoing coronary vascular surgery (26, 27). Coronary sinus oxygen content increased in parallel with coronary sinus blood flow, while myocardial oxygen corisumption and left ventricular stroke work remained unaffected. Adenosine hypotension is therefore associated with a hyperkinetic coronary circulation where the flow increase is unrelated to myocardial metabolic demand.

There were no signs of coronary vasodilation during SNP hypotension; on the contrary, SNP reduced coronary sinus flow and did not affect coronary sinus oxygen content. The coronary flow data were in agreement with the SNP response in awake patients (10). The present reduction in myocardial flow was thus coupled to the reductions in left ventricular stroke work and myocardial oxygen consumption. The absence of reduced coronary vascular resistance in parallel with reduced systemic vascular resistance by SNP, therefore suggests smaller coronary than systemic influence of this drug. Nevertheless, experimental and clinical studies have indicated intramyocardial flow redistribution during SNP administration (28, 29). This could imply that redistribution may occur even in the absence of increased total myocardial flow. There are several implications of adenosine-induced coronary vasodilation. This vasodilation may be beneficial when myocardial oxygen delivery is enhanced or detrimental when adenosine would cause negative intramyocardial redistribution of [low. Such coronary “steal” has been demonstrated during in tracoronary administration of adenosine in dogs with artificial coronary stenosis (3&32). There are also reports of myocardial ischaemia during adenosine-induced h) potension in patients with myocardial vascular disease (5). The patients in the present study were not evaluated with coronary angiography and therefore the incidence of coronary vascular disease is not known. The ECG recordings did not reveal any evidence of myocardial ischaemia during hypotension with either adenosine or SNP. Further, myocardial net release of lactate, which is a definite sign of ischaemia (33), was not observed in any case. However, coronary sinus lactate concentration increased in five of the six patients during adenosine infusion. The increase in coronary sinus lactate in these five patients may reflect an increased myocardial lactate formation caused by adenosine. Myocardial lactate production during adenosine infusion (30-60 pg * kg- . min - ’) has prrviously been demonstrated in patients after coronary artery by-pass grafting ( 3 4 ) . Coronary sinus lactate concentration may also be altered by the marked iricrease in myocardial blood flow, and thereby substrate (lactate) delivery, as suggested in canine studies during ATP and adenosine hypotension (35). In conclusion, a 25% reduction in mean arterial pressure induced by adenosine caused a hyperkinetic circulation in the myocardial vascular bed in tht-se patients with peripheral vascular disease. In contrast, there were no signs of hyperkinetic myocardial circulation during SNP-induced hypotension to the same blood pressure level. Adenosine acts as a powerlul coronary vasodilator and increases coronary sinus

’

A. oWA1.L AND A. SOLLEVI 220 t,lood flow, SNp reduces coronary blood nilroprusside and their control by cryptenaminr. dneslh .ha/,! 1980: 59: 909-916. llow in parallel with decreased myocardial work. 18. Conradsson T-B, Clarke B, Dixon C M S, llalton K N, Bartics I'

ACKNOWLEDGEMENTS 'I'his study was supported by grants from the Swedish Medical Research Council (Proj No 7485), the Swedish Association against Heart and Chrst Diseases. the Wiberg Foundation, and the Karolinska Institutr.

REFERENCES I . Sollrvi A, Lagerkranser M. lrestedt L. Gordon E, Lindqvi\t C. C m r o l k d hypotension with adenosinr in cerebral ane.ui'ysm surgery. ilnesthesiulogy 1984: 61: 400405. 2. Lagerkranser M, Gordon E,Rudrhill A. Cardiovascular effects of sodium nitroprusside in cerebral aneurysm surgery. Arlo h o e s //itsioI Scand 1980: 24: 426132. 3 . Marcus M L. Autoregulation in he coronary Circulation. In: The corollary circulation in health and disease New York: McGrawHill, 1983: 94-96. B A. Effects ofllalo4. Hickey R F. Sybrrt P E. Verrier E D, thane, rnfluranr, and isoflurane on coronary blood flow autorcgulation and coronary vascular reserve i n the canine heart. AneslhesiO I O ~ J 1988: I 68: 21-30, 5. Owall A, Gordon E, Lagerkranser M, Lindquist C, Rudehill A, Sollevi A. Clinical experience with adenosine for controlled hypo,-Inalp. 1987: 66: msiotl during cerebral a r l e v s m s v ~ r y:Ineslh . 229-234. 6. Owall A, Lavrkranser M, Sollevi A. Elfects ofadenosine-induced hypotension on myocardial hernodynamics and metabolism during cerebral aneurysm surgery. dnesth h a l f 1988: 67: 228-232. 7. A, JIrnberg p-0, Brodin L-n, solhi A. Effects ofadt'nosine-induced hypotension on myorardial hernodynamics and metabolism in fentanyl-anesthetized patients with peripheral vawular disease. Annfheszology 1988: 69: 4 16-42 I . 8. 'l'inker J H, Michenfelder J D. Sodium nitroprusside: pharmacol~lgy, toxicology and therapeutics. .4ntsthesioky 1976: 45: 340-354. 9. Makt;ibi M, Warner D,SokoU M et al. Comparison ofnitroprusside, nitroglycerin. and deep isoflurane anesthesia for induced hypotrnsion. .Wertrosurgery 1986: 19: 350-355. 10. k r n M J, M e n S D,Park R C, O'Rourke R A. Alteration.; irl rrgional myocardial blood flow aftrr nitroprusside and niiroglycerine in patients with and without significant coronar). artery disease. An! .7 Cnrdiol 1986: 58: 443448. I I . Vrsry C J , Cole P V, Simpson P J . Cyanide and thioryanate (011centratinns following sodium nitroprusside infusion in man. / { r J Anat.il/i 1976: 48: 651-660. I!?. Fahmy N R, Sunder N , MossJ , Slater E,Lappas D G. Tachypliylaxis to nitroprusside: role ofthe renin-angiotensin system and c'iterholamines in its development. ilnesthesiolo,g 1979: 51: S72. 13. Kliambatta H J , Stone J G , Khan E. Hypertension during a i m thesia on discontinuation of sodium nitroprusside-induced hypotension. .4nesthtstolo.cy 1979: 51: 127-130. 14. Khambatta H J , Stone J C , Khan E. Propranolol alters renin rrIrasv during nitroprusside-inducrd hypotrnsion and prevents Iiypertrnsion on discontinuation of nitroprusside. Anesth r l n a l ~I98 1 : 60: 569-573. 15. 'I'frlt-Hansen l', Siggaard-Andersen 0. Lartate and pyruvate tietermination in 50 p1 whole blood. Srand J Chi Lah Inorsf 1971: 27: 15-19. 16. Srttrrgren G. 'Ihe calculation ofleft ventricular stroke work indrx. .Irlu Anarsthesiol Scand 1986: 30: 450452. 17. Flackc J W, Flacke W E, Cant J W. Krllcx rrsponses to sodium-

J . Efferts ofadenosinr on automatic w n t r o l of heart ratr i n m i i n . .4cta PhysiolScand 1987: 131: 524-531. 19. Fuller R v, Maxwell D L, Conradsson 'I'-B, Dixon