- Email: [email protected]
Contractile, metabolic and arrhythmogenic effects of ionic and nonionic contrast agents in the isolated rat heart
Contractile, metabolic and arrhythmogenic effects of ionic and nonionic contrast agents in the isolated rat heart lntracoronary administration of contrast agents may be associated with contractile dysfunction and arrhythmias. To further establish the mechanisms of these alterations, we studied high-energy phosphate metabolism, developed pressure, the occurrence of arrhythmias, and the effects of verapamil during infusion of ionic and nonionic agents in isovolumic, retrogradely perfused rat hearts using 31P nuclear magnetic resonance imaging (NM). Diatriroate meglumine (Renografin) infusion reduced developed pressure (DP) to 17.1 + 3.4% (p < 0.001) of the control level, and immediately following termination of the infusion, sudden ventricular tachycardia (VT) was observed in four of six hearts. In the presence of verapamil, meglumine reduced DP to 13 + 1.9% of control values and none of these six hearts developed VT. lopamidol infusion in the presence of verapamil (n = 6) and alone (n = 6) resulted in a decrease in DP to 67% of control value, and no arrhythmias, significant change in high-energy phosphate levels, or changes in pH were observed. These results suggest that contrast-induced contractile depression is not mediated by changes in high-energy phosphate metabolism or pH. Arrhythmias associated with meglumine administration alone and suppressed by verapamil are probably related to calcium loading. (AM HEART J 1992;124:651.)
Sean T. Gloth, MD, Gary Gerstenblith,
MD, and Jeffrey A. Brinker,
MD
Baltimore, Md.
Hemodynamic and electrophysiologic changes accompany the intracoronary infusion of contrast agents in man and in animal mode1s.l The mechanisms responsible for these alterations are not completely understood. 1l 2 Although osmolality is high in contrast agents (1940mOsm/kgfor diatrizoate meglumine [Renografin, Squibb Diagnostics, New Brunswick, N.J.] and 796 mOsm/kg for iopamidol), this cannot be the only mechanism of contractile depression,13g4 because the intracoronary injection of 1800 mOsm/kg of mannitol produces a minimal decrease in left ventricular function in the canine heart. Myocardial depression is instead believed to be related to the ionic consequences of contrast,3Y5 possibly related to changes in intra- and extracellular calcium. From the Peter Belfer Laboratory of the Cardiology Division, Department of Medicine, The Johns Hopkins Hospital. Supported by the National Heart, Lung, and Blood Institute Specialized Center of Research (SCOR) Grant No. HL 17655-E Dr. Gloth was supported by a fellowship grant from the American Heart Association, Maryland Affiliate Inc. Received for publication Oct. 31, 1991; accepted March 6, 1992. Reprint requests: Sean T. Gloth, MD, Carnegie 591, Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21205. 4/l/38903
Sarcolemmal calcium exchange and content are both thought to be reduced.4 The hypotensive response to ionic agents is attenuated with the addition of calcium to contrast and is potentiated in the presence of calcium channel blocking agents.6j 7 The latter may in part be a result of peripheral vasodilation as well as a direct effect on the heart.8 In addition, it has been demonstrated that coronary sinus calcium falls to a greater extent with exposure to ionic contrast, which has a sodium concentration of 190 mEq/L, than with exposure to nonionic contrast, which has negligible sodium content, or to ionic contrast in the presence of verapamil.7 Furthermore, intracellular calcium is depressed in cultured ventricular cells by diatrizoate meglumine and this depression is reversed by the addition of calcium to the solution to account for chelation by diatrizoate meglumine.g Ventricular fibrillation is a major risk of coronary angiography and one of the most common lifethreatening comp1ications.l Hyperosmolality of the contrast has been cited as a cause of the increased fibrillatory propensity,1° and this effect is thought to be mediated via an increase in the QT, interval caused by changes in the transsarcolemmal Na/Ca gradient.l, 5, llq I2 Th e addition of calcium chloride 651
652
Glath, Gerstenblith, and Brinker
American
0
IONIC
NON-IONIC
1. Effect of contrast infusion on developed pressure (DP). Concentration of contrast was 15.25 vol !Y. Control DP was 112 + 3 mm Hg and 124 + 6 mm Hg in the ionic and nonionic group, respectively. The difference between the two groups was significant at p < 0.001.
Fig.
(CaC12) to diatrizoate meglumine or the use of high osmolal ionic contrast without the calcium binding additives of diatrizoate meglumine attenuates QT, prolongation and prevents spontaneous and induced ventricular tachycardia (VT).l, *a13,I4 Nonionic con-
trast is associated with less prolongation of the QT interval as well as adecreased incidence of VT33 lo, la, IF, compared with high osmolal ionic contrast agents. Ib would appear that the low osmolal ionic contrast (ioxaglate) occupies a position between the ionic and
the nonionic agents. l6 The mechanisms responsible for the development of the ventricular arrhythmias have not been elucidated.l, l1 The extent of metabolic alterations induced by contrast agents is more difficult to study and is thus poorly understood, Metrizamide, with its glucose side chain, inhibits hexokinase and causes neurotoxicity
as a result.17 Diatrizoate meglumine increases myocardial blood flow, at least in part as a result of metabolic actions.i8 Diatrizoate meglumine also causes increased uptake of free fatty acid and a decreased uptake of glucose. lg, 2oA study in patients with cor-
onary artery disease has also demonstrated that diatrizoate meglumine infusion results in an increase in coronary sinus lactate because of increased lactate
production. Although this was ascribed to an “ischemic” effect of the agent, it could not be determined whether the result was mediated via an effect on the coronary vasculature or directly on myocardial metabolism.20 31P nuclear magnetic resonance (NMR) spectroscopy offers an opportunity to study isolated cardiac metabolism and to define the direct effects of contrast on myocardial high-energy phosphate metabolism and pH. The purpose of this study was to examine, in the isolated intact heart, whether
September 1992 Heart Journal
contrast media-induced changes in pH and high-energy phosphate metabolism accompany the depression of left ventricular developed pressure. In addition, the role of calcium overload as a mechanism for the development of ventricular arrhythmias and depressed contractility was examined. NMR spectroscopy offers an excellent means of studying cardiac metabolism because it permits simultaneous measurements of pH, phosphocreatine, adenosine triphosphate (ATP), and left ventricular (LV) function in the intact heart. The role of calcium overload was studied indirectly by verapamil infusion. Diatrizoate meglumine was selected as the ionic agent because it is the most commonly used traditional hyperosmolal agent and is considered prototypical of this group despite the fact that, is formulated with calcium binding additives. As nonionic agents are associated with less toxicity than ionic agents, iopamidol was also studied. METHODS Heart preparation.
Retired male breeder rats, weighing 500 to 600 gm, were heparinized with 1000U of heparin sodium and anesthetized with 50 to 100mg of pentobarbital sodium, both via intraperitoneal administration. Hearts were rapidly excisedand retrogradely perfused through the cannulated aorta at a constant flow of 15 cm”/min, controlled by a Masterflex 756210 perist,altic pump (ColeParmer Instrument Co., Chicago, Ill.). The hearts were paced with bipolar contact electrodes at 75 beats/min after surgical atrioventricular (AV) block. A latex balloon attached to the end of 190polyethylene tubing was inserted into the left ventricle and was filled to a volume resulting in an end-diastolic pressureof 8 to 12mm Hg. The tubing wasconnectedto a Statham P23Db pressuretransducer (Viggo Spectramed Inc., Critical Care Division, Oxnard, Calif.) and left ventricular pressureswere continuously recorded on a Gould recorder (Gould Inc., Test & Measurement Recording Systems Division, Valley View, Ohio). Perfusates. The hearts were perfused with modified HEPES-buffered perfusate at 37” C at a pH of 7.4 containing (in millimoles per liter) sodium chloride, 140; potassium chloride, 5; calcium chloride, 1.5; HEPES, 6; magnesiumchloride, 1.2;and glucose,15.The solution also contained lidocaine, 5 mg/L, and insulin, 5 IJ/L. The perfusate in the slP NMR protocols also contained 2-deoxyglucose,5 mmol/L. Verapamil wasinfused at a concentration of 125Kg/L and the contrast agentsdiatrizoate meglumine (Renografin-76) and iopamidol (Isovue-370,BristolMyers, Squibb) wereinfused at a concentration of 15.25’; . 31P nuclear magnetic resonance imaging. In the ‘jlP NMR protocols the perfused heartswere placedinto the 25 mm bore of a Bruker superconducting magnet (Bruker Medical Instruments, Inc., Billerica, Mass.) of 4.2 T field strength. The phosphorus resonanceis 72.89 MHz. Detailed NMR methods are described elsewhere.“’ In brief.
Volume
124
Number
3
n1P NMR spectra were obtained with a Bruker WH 180 spectrometer interfaced with a Nicolet 1080 computer (Nicolet Instruments, Madison, Wise.). Pulsed, Fouriertransformed minimally saturated spectra were obtained with a 25 msec pulse and a Z-second period and changes in tissue contents of ATP, phosphocreatine (PCR), and inorganic phosphate are derived from the area under the individual peaks. Previous studieszl have shown that changes in ATP, phosphocreatine, and inorganic phosphate measured hy these NMR techniques were validated by biochemical determinations. Changes in pH were determined using 2-deoxyglucose, as the inorganic phosphate peak is generally too small under normal perfusion conditions. Temporary perfusion with 2-deoxyglucose for 15 to 30 minutes resulted in an intracellular, stable deoxyglucose6-phosphate peak2” that was used to measure intracellular pH. This concentration of deoxyglucose does not alter contractile performance or high-energy phosphate levels.2” Protocol. A total of 24 hearts were studied. All hearts were initially shownto have stable contractile function for at least 20 minutes. In the first set of experiments, the high-energy phosphate protocol, 12 hearts were infused with the 2-deoxyglucosefor 15 to 25 minutes. The hearts were then infused through a side port with either an ionic (n = 6) or a nonionic (n = 6) contrast agent for 20 minutes and were then perfused with contrast-free solution. In the secondset of experiments, additional hearts (n = 12) were perfused with verapamil, 125pg/ml for 20 minutes following the 20-minute stabilization period. These hearts were then infused with either the ionic (n = 6) or the nonionic (n = 6) contrast agent for 20 minutes. Data analysis. The data are expressedasmean t SEM. The results are analyzed usingpaired and unpaired t tests. A p value of SO.05 was consideredsignificant. RESULTS Contractile
function. Verapamil, ionic contrast, and nonionic contrast all independently depress contractile function. As illustrated in Fig. 1, diatrizoate meglumine (Renografin) and iopamidol, infused at a concentration of 15.25 ~01%) caused a decline in developed pressure (DP) t,hat stabilized by 20 minutes at 17 + 3.4% and 86.7 f 3.2% of control, respectively (p < 0.001). As shown in Fig. 2, infusion of verapamil decreased DP to 44 ? 3.9 9’, of control (n = 12) (p < 0.001). Fig. 3 illustrates the effect of contrast on DP in the presence of verapamil. The addition of diatrizoate meglumine (Renografin) to the verapamil infusion resulted in a further decrease of the DP to 13 t l.gGr of control values. However, the addition of iopamidol actually increased DP to 87.0 f 7.9 % of control values (p < 0.001). Although the DP measurements above represent steady-state levels, in all experiments the DP had fallen to a greater extent than steady-state levels by 20 seconds. Metabolism. Representative examples of the NMR spectra with simultaneous DP tracings are shown in
Radiocontrast
on myocardial
metabolism/mechanics
CONTROL
VERAPAMIL
653
Fig. 2. Effect of verapamil on developed pressure(DP).
Concentration of verapamil was125 pg/L. Control DP was 127 + 6 mm Hg and 56 i 4 mm Hg in the control and verapamil groups, respectively. The difference in the two groups wassignificant at p < 0.001.
Fig. 3. Effect of contrast on developed pressure(DP) in the presenceof verapamil. BaselineDP was 118 t- 6 mm Hg before exposure to contrast and fell to 15 + 3 mm Hg after contrast was infused. Baseline DP was 115 + 7 mm Hg before exposureto nonionic contrast and fell to 87 ? 7 mm Hg after the agent was infused. The difference in the two groups was significant at p < 0.001.
Fig. 4. Note that there is no change in the ATP and PCr peaks. The summed results of ATP, PCr, and pH for diatrizoate meglumine (Renografin) and iopamido1 are presented in Table I and show as well that these parameters are not changed by contrast infusion. As demonstrated by 31PNMR spectroscopy, the contractile depression induced by the ionic and nonionic contrast agents is not mediated by changes in cardiac metabolism as assessedby high-energy phosphate levels and cellular pH. Arrhythmias. Ventricular tachyarrhythmias were seen only in the diatrizoate meglumine (Renografin) group. During constant infusion of ionic and nonionic contrast agents, none of the hearts developed ventricular tachycardia (VT). Upon restoration of nor-
654
Gloth, Gerstenblith,
and Brinker
American
NON-IONIC
CONTROL
September 1992 Heart Journal
INFUSION
29
I;
_i’ili_. P ATP
'
,
IO
”
.
,
0
,
.
-10
(
-20
PPM
CONTROL
0
-10
IO
-10
0
IONIC
t _7-
ii
u
-20
PPM
INFUSION
-20
PPM
Fig. 4. Representative examplesof the NMR spectra with simultaneousdeveloped pressuretracings (inset), during the control period (left panel), and during contrast infusion (right panel). The top half of the figure depicts the results in an experiment during infusion of nonionic contrast and the bottom half shows
the results during ionic contrast infusion. There is no changein slP NMR spectra with either agent. The developed pressurewas markedly depressedby ionic contrast but not significantly affected by nonionic contrast.
ma1 perfusate, however, four of six hearts previously infused with ionic conkast alone developed VT. However, at the termination of ionic contrast infusion in the presence of verapamil, none of the hearts developed arrhythmias. At the termination of perfusion with nonionic agents, none of the hearts developed ventricular arrhythmias. DISCUSSION Contractile
function. Exposure to ionic contrast depresses contractile function to a greater extent than exposure to nonionic contrast under normal perfusion conditions as well as in the presence of verapamil. The initial sharp fall in DP at 20 seconds supports extrapolation of these steady-state levels (required to obtain NMR spectra) to clinical situations. It is interesting to note that whereas in the presence of verapamil the addition of ionic contrast further depresses contractile function, the addition
of nonionic contrast enhances contractile function, and that DP returns to the same level present during nonionic contrast infusion without verapamil. The apparent negation of the verapamil-induced contractile depression by iopamidol may be a result of competitive inhibition at the same receptor. Metabolism. While it has been suggested that ionic as opposed to nonionic contrast causesan “ischemic” response,2o the present study demonstrates no significant change in cellular pH or high-energy phosphate levels as a result of contrast infusion. This would suggest that the depressant effect on left ventricular function is not caused by the effect of contrast on high-energy phosphate metabolism. The discrepancy between these findings and those of Wisneski et a1.16may be related to indirect effects of the ionic contrast agent in patients with obstructive coronary disease. Arrhythmias. The literature reports that ionic con-
Volume Number
124 3
trast administration increases the likelihood of ventricular arrhythmias. In this study a stable paced rhythm in the rat heart was maintained throughout exposure to 15.25 ~01% of diatrizoate meglumine (Renografin) and VT developed at the time of withdrawal of the agent. A prolongation of the QT, interval has been implicated in the mechanism of arrhythmia induction. l2 This probably reflects intracellular hypocalcemia during contrast infusion, as has been demonstrated in cultured ventricular cells using indo-1.g We postulate that the arrhythmia may actually be induced as a result of rapid transsarcolemmal calcium influx upon the cessation of contrast infusion. In red blood cells and cardiac Purkinje cells, elevated intracellular calcium increases the potassium conductancez4 and it is possible that abrupt changes in potassium are responsible for the VT in our experiments. The mechanism for arrhythmia development, rapid transsarcolemmal calcium influx, would explain the observation that bradycardia exerts a protective effect on the incidence of VT as a consequence of delayed washout of the contrast,l which may result in a gradual, rather than rapid, calcium influx. Although it was not possible to obtain direct measures of intracellular calcium in this model, the hypothesis that there is rapid transsarcolemmal calcium influx on a return to normal perfusion conditions is also supported by the marked overshoot in developed pressure at the termination of contrast infusion. It is also supported by the fact that verapamil, which blocks calcium influx, prevents the development of VT, and by the finding of a previous report that the addition of calcium to ionic contrast agents, known to chelate calcium, results in both a decrease of intracellular loss9 and of the propensity for ventricular tachyarrhythmias.’ The addition of verapamil, which blocks calcium influx, appears to prevent the development of VT. Thus both the addition of calcium and verapamil to ionic contrast prevents the development of VT by preventing rapid transsarcolemmal calcium influx. The decrease in extracellular calcium because of chelation associated with ionic dye is the initiating event that is responsible for transsarcolemmal calcium influx when the contrast is removed. The addition of calcium diminishes the fall in extracellular calcium in the presence of contrast. The addition of verapamil decreases cellular calcium gain when contrast is removed. Nonionic contrast, which has less of an effect on intracellular calcium,l is associated with a lower incidence of ventricular tachyarrhythmias. Conclusions. Although both ionic and nonionic contrast agents depress myocardial function, the effect is greater with an ionic agent. Depression is not
Radiocontrast
on myocardial
metabolism/mechanics
655
Table I. Effect of contrast infusion on adenosine triphos-
phate (ATP), phosphocreatine (PCr), and pH Diatrizoate meglumine
ATP* PCr* PHt
100.35 102.62 7.07
Iopamidol
97.46 107.86 7.05
*ATP and PCr are expressed as a percent of the baseline value before administration of the contrast agent. tBaseline pH was 7.06 + 0.03 in t,he diatrizoate meglumine group and 7.03 + 0.03 in the iopamidol group.
the result of changes in high-energy metabolism, but is probably a consequence of changes in intracellular calcium. This mechanism may underlie the clinical differences between diatrizoate meglumine (Renografin) and other ionic agents that are not formulated with calcium binders. Arrhythmias associated with diatrizoate meglumine (Renografin) administration occur at the termination of contrast infusion and are probably related to calcium loading at that time. It is likely that these agents act in part via an effect at the calcium receptor, since verapamil eliminates diatrizoate meglumine (Renografin) withdrawal arrhythmias and iopamidol negates verapamil’s negative inotropic actions. REFERENCES
1. Murdock DK, Euler DE, Becker DM, Murdock JD, Scanlon PJ, Gunnar RM. Ventricular fibrillation during coronary angiography: an analysis of mechanisms. AM HEART J 1985; 10926573. 2 Dawson P. Chemotoxicity of contrast media and clinical adverse effects: a review. Invest Radio1 1985;2O(suppl):S84-91. 3. Gertz EW, Wisneski JA, Chiu D, Akin JR, Hu C. Clinical superiority of a new nonionic contrast agent (iopamidol) for cardiac angiography. J Am Co11 Cardiol 1985;5:250-8. 4. Pensabene J, Rosenblum J, Klein LW. Physiologic effects of contrast media used in coronary angiography. J Invasive Cardiol 1990;2:21-35. 5. Murdock DK, Piao ZE, Euler DE, et al. The use of programmed electrical stimulation to assess the fibrillatory propensity of ionic and nonionic contrast media. Invest Radio1 1985;20:579-82. 6. Morris DL, Wisneski JA, Gertz EW, Wexman M, Axelrod R, Langberg JJ. Potentiation by nifedipine and diltiazem of the hypotensive response after contrast angiography. J Am Co11 Cardiol 1985;6:785-91. 7. Higgins CB, Kuber M, Slutsky RA. Interaction between verapamil and contrast media in coronary arteriography: comparison of standard ionic and new nonionic media. Circulation 1983;68:628-35. 8. Chahine RA, Raizner AE. The mechanism of hvnotension following angiography. Invest Radio1 1976;11:472-8. 9. Bell DA, Peeters GA, Davis WL, Kohmoto 0, Nelson JA, Barry WH. Effects of contrast media on calcium transients and motion in cultured ventricular cells. Invest Radio1 1988;23:842-6. 10. Piao ZE, Murdock DK, Hwang MH, Raymond RM, Scanlon PJ. Contrast media-induced ventricular fibrillation. Invest Radio1 1988;23:466-70. 11. Popio KA, Ross AM, Oravec JM, Ingram JT. Identification
656
12. 13. 14. 15. 16.
17.
18.
Gloth, Gerstenblith, and Brinker
and description of separate mechanisms for two components of Renografin cardiotoxicity. Circulation 1978;58:520-8. Wolf GL, Hirshfeld JW. Changes in QT, interval induced with Renografin-76 and hypaque-76 during coronary arteriography. J Am Co11 Cardiol 1983;6:1489-92. Thomson KR, Violante MR, Kenyon T, Fischer HW. Reduction in ventricular fibrillation using calcium-enriched Renografin 76. Invest Radio1 1978;13:238-40. Fischer HW, Thomson KR. Contrast media in coronary arteriography: a review. Invest Radio1 1978;13:450-9. Wolf GL, Mulry CS, Laski PA, Kilzer K. Changes in ventricular fibrillation threshold induced by contrast agents during acute coronary artery occlusion. Invest Radio1 1983;18:145-8. Wisneski JA, Gertz EW, Dahlgren M, Muslin A. Comparison of low osmolality ionic (ioxaglate) versus nonionic (iopamidol) contrast media in cardiac angiography. Am J Cardiol 1989; 63:489-g& Simon JH, Ekholm SE, Morris TW, Fonte DJ. Further support for the glucose hypothesis of metrizamide toxicity: the effect of metrizamide and glucose analogue-free contrast media on hexokinase. Invest Radio1 1987;22:137-40. Gerber KH, Higgins CB. Comparative effects of ionic and
American
19. 20. 21.
22. 23.
September 1992 Heart Journal
nonionic contrast materials on coronary and peripheral blood flow. Invest Radio1 1982;17:292-8. Wiesneski JA, Gertz EW, Neese R, et al. Myocardial metabolic alterations after contrast angiography. Am J Cardiol 1982: 50:239-45. Wisneski JA, Gertz EW, Neese RA, Morris DL. Absence of myocardial biochemical toxicity with a nonionic contrast agent (iopamidol). AM HEART J 1985;110:609-17. Hoerter JA, Miceli MV, Renlund DG, Jacobus WE, Gerstenblith G, Lakatta EG. A phosphorus-31 nuclear magnetic resonance study of the metabolic, contractile, and ionic consequences of induced calcium alterations in the isovolumic rat heart. Circ Res 1986;58:539-51. Bailey IA, Williams SR, Radda GK, Gadian DG. Activity of phosphorylase in total global ischaemia in the rat heart. Biothem d 1981;196:171-8. Allen DG, Eisner DA, Morris PG, Pirolo JS, Smith GL. Metabolic consequences of increasing intracellular calcium and force production in perfused ferret hearts. J Physiol 1986;376:121-41.
24. Isenberg G. Is potassium conductance of cardiac Purkinje hbres controlled by [Cast];? Nature 1975;253:273-4.
Sean T. Gloth, MD, Gary Gerstenblith,
MD, and Jeffrey A. Brinker,
MD
Baltimore, Md.
Hemodynamic and electrophysiologic changes accompany the intracoronary infusion of contrast agents in man and in animal mode1s.l The mechanisms responsible for these alterations are not completely understood. 1l 2 Although osmolality is high in contrast agents (1940mOsm/kgfor diatrizoate meglumine [Renografin, Squibb Diagnostics, New Brunswick, N.J.] and 796 mOsm/kg for iopamidol), this cannot be the only mechanism of contractile depression,13g4 because the intracoronary injection of 1800 mOsm/kg of mannitol produces a minimal decrease in left ventricular function in the canine heart. Myocardial depression is instead believed to be related to the ionic consequences of contrast,3Y5 possibly related to changes in intra- and extracellular calcium. From the Peter Belfer Laboratory of the Cardiology Division, Department of Medicine, The Johns Hopkins Hospital. Supported by the National Heart, Lung, and Blood Institute Specialized Center of Research (SCOR) Grant No. HL 17655-E Dr. Gloth was supported by a fellowship grant from the American Heart Association, Maryland Affiliate Inc. Received for publication Oct. 31, 1991; accepted March 6, 1992. Reprint requests: Sean T. Gloth, MD, Carnegie 591, Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21205. 4/l/38903
Sarcolemmal calcium exchange and content are both thought to be reduced.4 The hypotensive response to ionic agents is attenuated with the addition of calcium to contrast and is potentiated in the presence of calcium channel blocking agents.6j 7 The latter may in part be a result of peripheral vasodilation as well as a direct effect on the heart.8 In addition, it has been demonstrated that coronary sinus calcium falls to a greater extent with exposure to ionic contrast, which has a sodium concentration of 190 mEq/L, than with exposure to nonionic contrast, which has negligible sodium content, or to ionic contrast in the presence of verapamil.7 Furthermore, intracellular calcium is depressed in cultured ventricular cells by diatrizoate meglumine and this depression is reversed by the addition of calcium to the solution to account for chelation by diatrizoate meglumine.g Ventricular fibrillation is a major risk of coronary angiography and one of the most common lifethreatening comp1ications.l Hyperosmolality of the contrast has been cited as a cause of the increased fibrillatory propensity,1° and this effect is thought to be mediated via an increase in the QT, interval caused by changes in the transsarcolemmal Na/Ca gradient.l, 5, llq I2 Th e addition of calcium chloride 651
652
Glath, Gerstenblith, and Brinker
American
0
IONIC
NON-IONIC
1. Effect of contrast infusion on developed pressure (DP). Concentration of contrast was 15.25 vol !Y. Control DP was 112 + 3 mm Hg and 124 + 6 mm Hg in the ionic and nonionic group, respectively. The difference between the two groups was significant at p < 0.001.
Fig.
(CaC12) to diatrizoate meglumine or the use of high osmolal ionic contrast without the calcium binding additives of diatrizoate meglumine attenuates QT, prolongation and prevents spontaneous and induced ventricular tachycardia (VT).l, *a13,I4 Nonionic con-
trast is associated with less prolongation of the QT interval as well as adecreased incidence of VT33 lo, la, IF, compared with high osmolal ionic contrast agents. Ib would appear that the low osmolal ionic contrast (ioxaglate) occupies a position between the ionic and
the nonionic agents. l6 The mechanisms responsible for the development of the ventricular arrhythmias have not been elucidated.l, l1 The extent of metabolic alterations induced by contrast agents is more difficult to study and is thus poorly understood, Metrizamide, with its glucose side chain, inhibits hexokinase and causes neurotoxicity
as a result.17 Diatrizoate meglumine increases myocardial blood flow, at least in part as a result of metabolic actions.i8 Diatrizoate meglumine also causes increased uptake of free fatty acid and a decreased uptake of glucose. lg, 2oA study in patients with cor-
onary artery disease has also demonstrated that diatrizoate meglumine infusion results in an increase in coronary sinus lactate because of increased lactate
production. Although this was ascribed to an “ischemic” effect of the agent, it could not be determined whether the result was mediated via an effect on the coronary vasculature or directly on myocardial metabolism.20 31P nuclear magnetic resonance (NMR) spectroscopy offers an opportunity to study isolated cardiac metabolism and to define the direct effects of contrast on myocardial high-energy phosphate metabolism and pH. The purpose of this study was to examine, in the isolated intact heart, whether
September 1992 Heart Journal
contrast media-induced changes in pH and high-energy phosphate metabolism accompany the depression of left ventricular developed pressure. In addition, the role of calcium overload as a mechanism for the development of ventricular arrhythmias and depressed contractility was examined. NMR spectroscopy offers an excellent means of studying cardiac metabolism because it permits simultaneous measurements of pH, phosphocreatine, adenosine triphosphate (ATP), and left ventricular (LV) function in the intact heart. The role of calcium overload was studied indirectly by verapamil infusion. Diatrizoate meglumine was selected as the ionic agent because it is the most commonly used traditional hyperosmolal agent and is considered prototypical of this group despite the fact that, is formulated with calcium binding additives. As nonionic agents are associated with less toxicity than ionic agents, iopamidol was also studied. METHODS Heart preparation.
Retired male breeder rats, weighing 500 to 600 gm, were heparinized with 1000U of heparin sodium and anesthetized with 50 to 100mg of pentobarbital sodium, both via intraperitoneal administration. Hearts were rapidly excisedand retrogradely perfused through the cannulated aorta at a constant flow of 15 cm”/min, controlled by a Masterflex 756210 perist,altic pump (ColeParmer Instrument Co., Chicago, Ill.). The hearts were paced with bipolar contact electrodes at 75 beats/min after surgical atrioventricular (AV) block. A latex balloon attached to the end of 190polyethylene tubing was inserted into the left ventricle and was filled to a volume resulting in an end-diastolic pressureof 8 to 12mm Hg. The tubing wasconnectedto a Statham P23Db pressuretransducer (Viggo Spectramed Inc., Critical Care Division, Oxnard, Calif.) and left ventricular pressureswere continuously recorded on a Gould recorder (Gould Inc., Test & Measurement Recording Systems Division, Valley View, Ohio). Perfusates. The hearts were perfused with modified HEPES-buffered perfusate at 37” C at a pH of 7.4 containing (in millimoles per liter) sodium chloride, 140; potassium chloride, 5; calcium chloride, 1.5; HEPES, 6; magnesiumchloride, 1.2;and glucose,15.The solution also contained lidocaine, 5 mg/L, and insulin, 5 IJ/L. The perfusate in the slP NMR protocols also contained 2-deoxyglucose,5 mmol/L. Verapamil wasinfused at a concentration of 125Kg/L and the contrast agentsdiatrizoate meglumine (Renografin-76) and iopamidol (Isovue-370,BristolMyers, Squibb) wereinfused at a concentration of 15.25’; . 31P nuclear magnetic resonance imaging. In the ‘jlP NMR protocols the perfused heartswere placedinto the 25 mm bore of a Bruker superconducting magnet (Bruker Medical Instruments, Inc., Billerica, Mass.) of 4.2 T field strength. The phosphorus resonanceis 72.89 MHz. Detailed NMR methods are described elsewhere.“’ In brief.
Volume
124
Number
3
n1P NMR spectra were obtained with a Bruker WH 180 spectrometer interfaced with a Nicolet 1080 computer (Nicolet Instruments, Madison, Wise.). Pulsed, Fouriertransformed minimally saturated spectra were obtained with a 25 msec pulse and a Z-second period and changes in tissue contents of ATP, phosphocreatine (PCR), and inorganic phosphate are derived from the area under the individual peaks. Previous studieszl have shown that changes in ATP, phosphocreatine, and inorganic phosphate measured hy these NMR techniques were validated by biochemical determinations. Changes in pH were determined using 2-deoxyglucose, as the inorganic phosphate peak is generally too small under normal perfusion conditions. Temporary perfusion with 2-deoxyglucose for 15 to 30 minutes resulted in an intracellular, stable deoxyglucose6-phosphate peak2” that was used to measure intracellular pH. This concentration of deoxyglucose does not alter contractile performance or high-energy phosphate levels.2” Protocol. A total of 24 hearts were studied. All hearts were initially shownto have stable contractile function for at least 20 minutes. In the first set of experiments, the high-energy phosphate protocol, 12 hearts were infused with the 2-deoxyglucosefor 15 to 25 minutes. The hearts were then infused through a side port with either an ionic (n = 6) or a nonionic (n = 6) contrast agent for 20 minutes and were then perfused with contrast-free solution. In the secondset of experiments, additional hearts (n = 12) were perfused with verapamil, 125pg/ml for 20 minutes following the 20-minute stabilization period. These hearts were then infused with either the ionic (n = 6) or the nonionic (n = 6) contrast agent for 20 minutes. Data analysis. The data are expressedasmean t SEM. The results are analyzed usingpaired and unpaired t tests. A p value of SO.05 was consideredsignificant. RESULTS Contractile
function. Verapamil, ionic contrast, and nonionic contrast all independently depress contractile function. As illustrated in Fig. 1, diatrizoate meglumine (Renografin) and iopamidol, infused at a concentration of 15.25 ~01%) caused a decline in developed pressure (DP) t,hat stabilized by 20 minutes at 17 + 3.4% and 86.7 f 3.2% of control, respectively (p < 0.001). As shown in Fig. 2, infusion of verapamil decreased DP to 44 ? 3.9 9’, of control (n = 12) (p < 0.001). Fig. 3 illustrates the effect of contrast on DP in the presence of verapamil. The addition of diatrizoate meglumine (Renografin) to the verapamil infusion resulted in a further decrease of the DP to 13 t l.gGr of control values. However, the addition of iopamidol actually increased DP to 87.0 f 7.9 % of control values (p < 0.001). Although the DP measurements above represent steady-state levels, in all experiments the DP had fallen to a greater extent than steady-state levels by 20 seconds. Metabolism. Representative examples of the NMR spectra with simultaneous DP tracings are shown in
Radiocontrast
on myocardial
metabolism/mechanics
CONTROL
VERAPAMIL
653
Fig. 2. Effect of verapamil on developed pressure(DP).
Concentration of verapamil was125 pg/L. Control DP was 127 + 6 mm Hg and 56 i 4 mm Hg in the control and verapamil groups, respectively. The difference in the two groups wassignificant at p < 0.001.
Fig. 3. Effect of contrast on developed pressure(DP) in the presenceof verapamil. BaselineDP was 118 t- 6 mm Hg before exposure to contrast and fell to 15 + 3 mm Hg after contrast was infused. Baseline DP was 115 + 7 mm Hg before exposureto nonionic contrast and fell to 87 ? 7 mm Hg after the agent was infused. The difference in the two groups was significant at p < 0.001.
Fig. 4. Note that there is no change in the ATP and PCr peaks. The summed results of ATP, PCr, and pH for diatrizoate meglumine (Renografin) and iopamido1 are presented in Table I and show as well that these parameters are not changed by contrast infusion. As demonstrated by 31PNMR spectroscopy, the contractile depression induced by the ionic and nonionic contrast agents is not mediated by changes in cardiac metabolism as assessedby high-energy phosphate levels and cellular pH. Arrhythmias. Ventricular tachyarrhythmias were seen only in the diatrizoate meglumine (Renografin) group. During constant infusion of ionic and nonionic contrast agents, none of the hearts developed ventricular tachycardia (VT). Upon restoration of nor-
654
Gloth, Gerstenblith,
and Brinker
American
NON-IONIC
CONTROL
September 1992 Heart Journal
INFUSION
29
I;
_i’ili_. P ATP
'
,
IO
”
.
,
0
,
.
-10
(
-20
PPM
CONTROL
0
-10
IO
-10
0
IONIC
t _7-
ii
u
-20
PPM
INFUSION
-20
PPM
Fig. 4. Representative examplesof the NMR spectra with simultaneousdeveloped pressuretracings (inset), during the control period (left panel), and during contrast infusion (right panel). The top half of the figure depicts the results in an experiment during infusion of nonionic contrast and the bottom half shows
the results during ionic contrast infusion. There is no changein slP NMR spectra with either agent. The developed pressurewas markedly depressedby ionic contrast but not significantly affected by nonionic contrast.
ma1 perfusate, however, four of six hearts previously infused with ionic conkast alone developed VT. However, at the termination of ionic contrast infusion in the presence of verapamil, none of the hearts developed arrhythmias. At the termination of perfusion with nonionic agents, none of the hearts developed ventricular arrhythmias. DISCUSSION Contractile
function. Exposure to ionic contrast depresses contractile function to a greater extent than exposure to nonionic contrast under normal perfusion conditions as well as in the presence of verapamil. The initial sharp fall in DP at 20 seconds supports extrapolation of these steady-state levels (required to obtain NMR spectra) to clinical situations. It is interesting to note that whereas in the presence of verapamil the addition of ionic contrast further depresses contractile function, the addition
of nonionic contrast enhances contractile function, and that DP returns to the same level present during nonionic contrast infusion without verapamil. The apparent negation of the verapamil-induced contractile depression by iopamidol may be a result of competitive inhibition at the same receptor. Metabolism. While it has been suggested that ionic as opposed to nonionic contrast causesan “ischemic” response,2o the present study demonstrates no significant change in cellular pH or high-energy phosphate levels as a result of contrast infusion. This would suggest that the depressant effect on left ventricular function is not caused by the effect of contrast on high-energy phosphate metabolism. The discrepancy between these findings and those of Wisneski et a1.16may be related to indirect effects of the ionic contrast agent in patients with obstructive coronary disease. Arrhythmias. The literature reports that ionic con-
Volume Number
124 3
trast administration increases the likelihood of ventricular arrhythmias. In this study a stable paced rhythm in the rat heart was maintained throughout exposure to 15.25 ~01% of diatrizoate meglumine (Renografin) and VT developed at the time of withdrawal of the agent. A prolongation of the QT, interval has been implicated in the mechanism of arrhythmia induction. l2 This probably reflects intracellular hypocalcemia during contrast infusion, as has been demonstrated in cultured ventricular cells using indo-1.g We postulate that the arrhythmia may actually be induced as a result of rapid transsarcolemmal calcium influx upon the cessation of contrast infusion. In red blood cells and cardiac Purkinje cells, elevated intracellular calcium increases the potassium conductancez4 and it is possible that abrupt changes in potassium are responsible for the VT in our experiments. The mechanism for arrhythmia development, rapid transsarcolemmal calcium influx, would explain the observation that bradycardia exerts a protective effect on the incidence of VT as a consequence of delayed washout of the contrast,l which may result in a gradual, rather than rapid, calcium influx. Although it was not possible to obtain direct measures of intracellular calcium in this model, the hypothesis that there is rapid transsarcolemmal calcium influx on a return to normal perfusion conditions is also supported by the marked overshoot in developed pressure at the termination of contrast infusion. It is also supported by the fact that verapamil, which blocks calcium influx, prevents the development of VT, and by the finding of a previous report that the addition of calcium to ionic contrast agents, known to chelate calcium, results in both a decrease of intracellular loss9 and of the propensity for ventricular tachyarrhythmias.’ The addition of verapamil, which blocks calcium influx, appears to prevent the development of VT. Thus both the addition of calcium and verapamil to ionic contrast prevents the development of VT by preventing rapid transsarcolemmal calcium influx. The decrease in extracellular calcium because of chelation associated with ionic dye is the initiating event that is responsible for transsarcolemmal calcium influx when the contrast is removed. The addition of calcium diminishes the fall in extracellular calcium in the presence of contrast. The addition of verapamil decreases cellular calcium gain when contrast is removed. Nonionic contrast, which has less of an effect on intracellular calcium,l is associated with a lower incidence of ventricular tachyarrhythmias. Conclusions. Although both ionic and nonionic contrast agents depress myocardial function, the effect is greater with an ionic agent. Depression is not
Radiocontrast
on myocardial
metabolism/mechanics
655
Table I. Effect of contrast infusion on adenosine triphos-
phate (ATP), phosphocreatine (PCr), and pH Diatrizoate meglumine
ATP* PCr* PHt
100.35 102.62 7.07
Iopamidol
97.46 107.86 7.05
*ATP and PCr are expressed as a percent of the baseline value before administration of the contrast agent. tBaseline pH was 7.06 + 0.03 in t,he diatrizoate meglumine group and 7.03 + 0.03 in the iopamidol group.
the result of changes in high-energy metabolism, but is probably a consequence of changes in intracellular calcium. This mechanism may underlie the clinical differences between diatrizoate meglumine (Renografin) and other ionic agents that are not formulated with calcium binders. Arrhythmias associated with diatrizoate meglumine (Renografin) administration occur at the termination of contrast infusion and are probably related to calcium loading at that time. It is likely that these agents act in part via an effect at the calcium receptor, since verapamil eliminates diatrizoate meglumine (Renografin) withdrawal arrhythmias and iopamidol negates verapamil’s negative inotropic actions. REFERENCES
1. Murdock DK, Euler DE, Becker DM, Murdock JD, Scanlon PJ, Gunnar RM. Ventricular fibrillation during coronary angiography: an analysis of mechanisms. AM HEART J 1985; 10926573. 2 Dawson P. Chemotoxicity of contrast media and clinical adverse effects: a review. Invest Radio1 1985;2O(suppl):S84-91. 3. Gertz EW, Wisneski JA, Chiu D, Akin JR, Hu C. Clinical superiority of a new nonionic contrast agent (iopamidol) for cardiac angiography. J Am Co11 Cardiol 1985;5:250-8. 4. Pensabene J, Rosenblum J, Klein LW. Physiologic effects of contrast media used in coronary angiography. J Invasive Cardiol 1990;2:21-35. 5. Murdock DK, Piao ZE, Euler DE, et al. The use of programmed electrical stimulation to assess the fibrillatory propensity of ionic and nonionic contrast media. Invest Radio1 1985;20:579-82. 6. Morris DL, Wisneski JA, Gertz EW, Wexman M, Axelrod R, Langberg JJ. Potentiation by nifedipine and diltiazem of the hypotensive response after contrast angiography. J Am Co11 Cardiol 1985;6:785-91. 7. Higgins CB, Kuber M, Slutsky RA. Interaction between verapamil and contrast media in coronary arteriography: comparison of standard ionic and new nonionic media. Circulation 1983;68:628-35. 8. Chahine RA, Raizner AE. The mechanism of hvnotension following angiography. Invest Radio1 1976;11:472-8. 9. Bell DA, Peeters GA, Davis WL, Kohmoto 0, Nelson JA, Barry WH. Effects of contrast media on calcium transients and motion in cultured ventricular cells. Invest Radio1 1988;23:842-6. 10. Piao ZE, Murdock DK, Hwang MH, Raymond RM, Scanlon PJ. Contrast media-induced ventricular fibrillation. Invest Radio1 1988;23:466-70. 11. Popio KA, Ross AM, Oravec JM, Ingram JT. Identification
656
12. 13. 14. 15. 16.
17.
18.
Gloth, Gerstenblith, and Brinker
and description of separate mechanisms for two components of Renografin cardiotoxicity. Circulation 1978;58:520-8. Wolf GL, Hirshfeld JW. Changes in QT, interval induced with Renografin-76 and hypaque-76 during coronary arteriography. J Am Co11 Cardiol 1983;6:1489-92. Thomson KR, Violante MR, Kenyon T, Fischer HW. Reduction in ventricular fibrillation using calcium-enriched Renografin 76. Invest Radio1 1978;13:238-40. Fischer HW, Thomson KR. Contrast media in coronary arteriography: a review. Invest Radio1 1978;13:450-9. Wolf GL, Mulry CS, Laski PA, Kilzer K. Changes in ventricular fibrillation threshold induced by contrast agents during acute coronary artery occlusion. Invest Radio1 1983;18:145-8. Wisneski JA, Gertz EW, Dahlgren M, Muslin A. Comparison of low osmolality ionic (ioxaglate) versus nonionic (iopamidol) contrast media in cardiac angiography. Am J Cardiol 1989; 63:489-g& Simon JH, Ekholm SE, Morris TW, Fonte DJ. Further support for the glucose hypothesis of metrizamide toxicity: the effect of metrizamide and glucose analogue-free contrast media on hexokinase. Invest Radio1 1987;22:137-40. Gerber KH, Higgins CB. Comparative effects of ionic and
American
19. 20. 21.
22. 23.
September 1992 Heart Journal
nonionic contrast materials on coronary and peripheral blood flow. Invest Radio1 1982;17:292-8. Wiesneski JA, Gertz EW, Neese R, et al. Myocardial metabolic alterations after contrast angiography. Am J Cardiol 1982: 50:239-45. Wisneski JA, Gertz EW, Neese RA, Morris DL. Absence of myocardial biochemical toxicity with a nonionic contrast agent (iopamidol). AM HEART J 1985;110:609-17. Hoerter JA, Miceli MV, Renlund DG, Jacobus WE, Gerstenblith G, Lakatta EG. A phosphorus-31 nuclear magnetic resonance study of the metabolic, contractile, and ionic consequences of induced calcium alterations in the isovolumic rat heart. Circ Res 1986;58:539-51. Bailey IA, Williams SR, Radda GK, Gadian DG. Activity of phosphorylase in total global ischaemia in the rat heart. Biothem d 1981;196:171-8. Allen DG, Eisner DA, Morris PG, Pirolo JS, Smith GL. Metabolic consequences of increasing intracellular calcium and force production in perfused ferret hearts. J Physiol 1986;376:121-41.
24. Isenberg G. Is potassium conductance of cardiac Purkinje hbres controlled by [Cast];? Nature 1975;253:273-4.