- Email: [email protected]
Coronary artery spasm after abrupt withdrawal of nitroglycerin in rabbits
EXPERIMENTAL STUDIES
Coronary Artery Spasm After Abrupt Withdrawal of Nitroglycerin in Rabbits WILLIAM C. REEVES, MD, LAWRENCE COOK, MD, MARY ANN WOOD, BS, and LAWRENCE WHITESELL, BS
The potential for abrupt withdrawal of nitroglycerin (NTG) to create coronary artery spasm was assessed in New Zealand white rabbits. In the control setting, electrocardiograms were taken from 7 anesthetized rabbits. The administration of intravenous ergonovine did not provoke ST-segment shifts or arrhythmias. Two inches of topical NTG (2 % ) was applied 3 times daily to a shaved area on the back of each rabbit over a 6-week period. Forty hours afler abrupt withdrawal of NTG, intravenous ergonovtne and indomethacin were given. Six of 7 rabbits had electrocardiographic changes: ventricular tachycardia in 2, ventricular fibrillation in 1, and significant (1 mm or more) ST-segment shifts in 5 rabbits. Three rabbits died. Sixty-four hours after NTG administration the remaining 4 rabbits were reexamined. One had baseline electrocardiographic evidence of severe myocardial ischemia. Repeat
ergonovine and indomethacin testing in the others revealed ventricular tachycardia progressing to asystole in 1, premature ventricular complexes in 1, and ST-segment elevation in another. Two of the remaining 4 rabbits died. Eighty-two hours after NTG administration the remaining 2 rabbits were found dead in their cages. Nitroblue tetrarolium studies revealed extensive myocardial infarction in both animals. Additional studies were performed in 10 normal rabbits. Neither ergonovine nor indomethacin induced ST-segment shifts or arrhythmias in this control population. It is concluded that abrupt withdrawal of NTG and subsequent administration of ergonovine and indomethacin created myocardial tschemia, ventricular arrhythmias and myocardial infarctions in these rabbits. The cause of these abnormalities is probably coronary artery spasm. (Am J Cardtol 1965;55:1066-1069)
Occupational habituation to nitrates, including withdrawal hazards of angina pectoris, coronary artery spasm and sudden death, are well d0cumented.l However, little is known regarding the safety of withdrawal of long-acting nitroglycerin (NTG) in patients treated for angina pectoris or congestive heart failure. We developed an experimental rabbit model characterized by abrupt withdrawal from long-acting NTG. This report describes our experience with this model using parenteral coronary vasoconstrictors to induce latent coronary artery spasm during the withdrawal period.
Series 1: Documentation of nitroglycerin absorption: Seven rabbits were anesthetized with intravenous sodium pentobarbital (30 mg/kg). Arterial blood pressure was measured through the central ear artery. Intravenous NTG was administered through a marginal ear vein in order to lower the mean arterial blood pressure 20 mm Hg from its baseline value. An initial dose of 10 pg/kg/min was administered over 3 minutes. This dose was doubled every 3 minutes until the appropriate blood pressure decrease was noted. Two inches of NTG ointment (2%, Kremers-Urban Company) was applied to a shaved area (16 square inches) in the interscapular region of the back 3 times daily (at 0900, 1300 and 1700 hours). During the seventh day of this treatment the rabbits were reanesthetized and the intravenous NTG protocol was repeated to the same blood pressure endpoint. Electrocardiographic analysis: Before topical NTG application, the 7 New Zealand white rabbits were anesthetized with intravenous sodium pentobarbital(30 mg/kg) and placed supine on a surgical table. Electrocardiographic (ECG) leads were attached to each extremity and the midsternum with subcutaneous needles. Simultaneous recordings of lead II and a single V lead were obtained with a Hewlett Packard Electrograph recorded at 50 mm/s using regular standardization (10 mm/mV). A baseline ECG recording was obtained. After
Methods Male New Zealand white rabbits that weighed 2.5 to 4.0 kg were used in this investigation. From the Milton S. Hershey Medical Center, Pennsylvania State University, Department of Cardiology, Hershey, Pennsylvania. Manuscript received October 23, 1984; revised manuscript received December 26, 1984, accepted December 31.1984. Address for reprints: William C. Reeves, MD, Mount Sinai Medical Center, Cardiovascular Disease Section, 950 North 12th Street, Milwaukee, Wisconsin 53233.
April 1,1985
this procedure, intravenous ergonovine (0.15 mg/kg) was given over 10 seconds. Electrocardiographic tracings were obtained immediately after ergonovine administration and at 30 seconds, 1,2,3,4 and 5 minutes after administration of ergonovine. A second intravenous ergonovine dose (0.15 mg/kg) was administered and the electrocardiogram recorded as previously described. Two inches of topical NTG (2%) was applied to the shaved interscapular area of each rabbit 3 times daily. This was done for 42 consecutive days. The interscapular area was shaved once a week during this period. Nitroglycerin administration was abruptly discontinued and 40 hours after the last administration of NTG the rabbits were reanesthetized. They were again attached to the electrograph and an initial dose of ergonovine (0.15 mg/kg) was given with ECG monitoring as previously described. This was followed by indomethacin (25 mg/kg) given intravenously over 30 seconds. Electrocardiographic monitoring was performed as with ergonovine administration. The remaining rabbits were allowed to recover and 24 hours later they were reanesthetized and ergonovine and indomethacin were again administered and monitored electrographically as at 40 hours. The remaining rabbits were returned to their cagesto recover. Myocardial infarction detection: The hearts from 2 rabbits that were found dead in their cages 82 hours after the last NTG administration were stained with nitroblue tetrazolium.2 Three transverse slices of the left ventricle were obtained from each heart. These were incubated in nitroblue tetrazolium solution for 30 minutes at 37°C. Nitroblue tetrazolium solution was made by dissolving nitroblue tetrazolium powder in distilled water at a concentration of 5 mg/ml. Ten milliliters of this solution were added to 10 ml of Sorensen’s phosphate buffer (1 M), pH 7.4 and 80 ml of distilled water. Normal heart muscle is stained by this technique. Infarcted myocardium remains unstained. Series 2: A second baseline control study was performed on 10 male New Zealand white rabbits. The rabbits were anesthetized and attached to the electrograph as previously described. Three intravenous injections were used in the following order: ergonovine (0.15 mg/kg), indomethacin (25 mg/kg), and ergonovine (0.15 mg/kg). A 5-minute ECG monitoring period was observed after each drug administration as described in series 1. Statistical analysis: The amount of intravenous NTG required to drop systemic arterial blood pressure before and after topical NTG administration was compared by the Student t test. A p value <0.05 was considered statistically significant.
Results Series 1: Nitroglycerin absorption: Initially, an average of 51 f 20 pg/kg (mean f standard error of the mean) of intravenous NTG was required to decrease mean systemic arterial blood pressure 20 mm Hg. During the seventh day of chronic NTG administration
FIGURE 2. Rabbit 5. Electrocardiogram after administration of indomethacin at 40 hours after the last dose of nitroglycerin. Upper tracing, lead II shows 2 mm of ST-segment elevation; lower tracing, lead V shows 4 mm of ST-segment elevation.
THE AMERICAN JOURNAL OF CARDIOLOGY Volume 55
1067
FIGURE 1. Rabbit 1. Electrocardiogram after administration of indomethacin at 40 hours after the last dose of nitroglycerin. Upper tracing, lead II shows downsloping ST-segment depression (2 mm) and T-wave inversion: lower tredng, a simultaneous lead V shows T-wave inversion without significant ST depression.
a significantly greater (p <0.05) amount of intravenous NTG, 95 f 30 pg/kg, was required to achieve the same decrease in systemic arterial blood pressure. This documents NTG absorption and the development of tolerance. Electrocardiographic ana2ysis: Before topical NTG administration, no significant ST-segment shifts (1 mm or greater) were noted in any rabbit at the baseline tracing or after administration of either ergonovine dose. No arrhythmias were noted. Forty hours after the last nitroglycerin dose: Ergonovine induced ST-segment elevation in 2 rabbits. Indomethacin administration resulted in ST-segment depression in 1 rabbit (Fig. l), ST-segment elevation in 4 rabbits (Fig. 2) and ventricular tachycardia and fibrillation in 2 rabbits. One rabbit had no ECG changes with ergonovine and after indomethacin administration had an immediate respiratory arrest and death with no antecedent ECG changes. Sixty-four ministration:
hours after the last nitroglycerin
ad-
Ergonovine induced isolated premature ventricular complexes in 1 rabbit and ST-segment elevation in another. Indomethacin provoked ventricular tachycardia that progressed to ventricular asystole and death in 1 rabbit. In 1 rabbit no drugs were administered because the baseline electrocardiogram exhibited 4 mm of ST-segment elevation with a reduced QRS complex in lead V and upsloping ST-segment depression in lead II (Fig. 3). Death followed shortly thereafter. Myocardial infarction detection: Transverse slices of the left ventricle from each rabbit found dead in its
1066
CORONARY
ARTERY SPASM AFTER NITROGLYCERIN WITHDRAWAL
FIGURE 3. Rabbit 5. Baseline electrocardiogram without drug provocation at 64 hours after t@ last nitroglycerin dose. Upper tracing, lead II Shows upsloping ST depression; lower tracing, lead V shows 4 mm of ST-segment elevation.
cage 82 hours after the last NTG administration had extensive myocardial infarctions (Fig. 4). Series 2: No ECG ST-segment changes or arrhythmias were observed after either ergonovine or indomethacin administrations. Discussion Sudden withdrawal after intensive occupational exposure to NTG can produce coronary artery spasm, myocardial infarction and sudden death.’ Development of similar complications caused by sudden withdrawal of long-acting NTG in patients with angina pectoris or congestive heart failure has not been well documented in published reports. Franciosa et als described 2 patients with coronary artery disease who died suddenly after discontinuing oral isosorbide dinitrate. Muller and Gunthe+ described a case of acute myocardial infarction occurring soon after withdrawal of NTG in a patient with variant angina. In clinical practice it would be difficult to establish such a causal relation between these complications and NTG withdrawal. A clinical study to evaluate the role of NTG withdrawal would be unethical. Furthermore, patients with coronary artery disease have coronary artery spasm and acute myocardial infarction and die suddenly independent of the use and withdrawal of NTG. The present study is an attempt to evaluate the
FIGURE 4. Transverse slice of the heart from a rabbit found dead in his cage 62 hours after administration of the last nitroglycerin dose. The nltroblue tetrazolium stain of the left ventricle at the papillary muscle level exhibits myocardlal Infarction (light areas), particularly evident in the ventricular septum.
role of NTG withdrawal in the development of such complications in an animal model. Caution must be used in attempting to directly extrapolate the results of our study in anesthetized rabbits to conscious humans. We selected rabbits for this investigation because they have been used successfully to demonstrate tolerance to NTG,5p6 a condition noted to occur in human vascular smooth muscle. Additionally, Rinzler et al7 developed a protocol using intravenous ergonovine and ECG monitoring to detect latent coronary artery disease in anesthetized rabbits. Ergonovine and indomethacin were chosen to test for latent coronary artery spasm in our experimental model. Ergonovine is a potent smooth-muscle vasoconstrictor commonly used to detect coronary artery spasm in human clinical studies.8 Indomethacin is a nonsteroidal anti-’ inflammatory agent. Its role in diminishing coronary flow is controversial9 However, it has been suggested to be a coronary artery vasoconstrictor, presumably through blockade of prostaglandin synthesis.1° Significant electrocardiographic ST-segment changes and ventricular arrhythmias after ergonovine and indomethacin were found only during the phase of NTG withdrawal. Six of 7 rabbits showed significant changes at the first testing interval 40 hours after the last dose of topical NTG. One rabbit had significant ST elevation after ergonovine, and 3 rabbits after indomethacin. One rabbit had significant ST depression after a 30-second episode of ventricular tachycardia after indomethacinj Another had ventricular tachycardia and ventricular fibrillation after indomethacin. At 64 hours after the last NTG administration, ST-segment shifts and ventricular arrhythmias could still be provoked with ergonovine and indomethacin. In addition, 1 rabbit had ECG evidence of severe myocardial ischemia during a baseline tracing. Only 2 rabbits survived this testing period and both were found dead in their cages the next morning. Nitroblue tetrazolium, a technique that allows macroscopic identification of early myocardial infarction by dehydrogenase alterations,2 revealed myocardial infarction in each instance. It is possible that platelet aggregation or a direct drug effect caused or contributed to the observed ahnormalities. Furthermore, we did not document histologically the absence of coronary atherosclerosis in our rabbits. However, coronary atherosclerosis probably did not play a role in the development of myocardial ischemia and infarction in our study, because rabbits are not known to spontaneously have this condition.ll We used large doses of ergonovine and indomethacin in this study. However, we do not believe that these doses created false-positive results, because neither agent provoked significant ECG abnormalities in the
April I,1985
control settings. Although our ergonovine dose (0.15 mg/kg) was larger than the usual dose administered in Rinzler’s7 series (0.05 mg/kg), it was well below the maximal dose used by him (1.0 mg/kg). The large indomethacin dose was chosen to ensure a pharmacologic effect in this animal species. Although we could not directly detect the presence of coronary artery spasm in our experimental model, the inducible ST-segment shifts, ventricular tachycardia and myocardial infarctions are suggestive of coronary artery spasm. In the setting of severe spasm, transmural ischemia results in ST-segment elevation, whereas lesser degrees of ischemia may cause ST-segment depression.12J3 Malignant ventricular rhythm disturbances are also found in humans with coronary artery spasm14 and may be induced by ergonovine in the clinical setting.15 We conclude that significant ST-segment shifts, malignant ventricular rhythm disturbances and myocardial infarction can be provoked in an experimental rabbit model after the administration of ergonovine and indomethacin during the period of immediate withdrawal from chronic NTG exposure. The probable cause of these abnormalities is coronary artery spasm. Acknowledgment: We acknowledge the expert secretarial assistance of Vicki Weisenburg and the editorial review of Dr.
Carl Christensenin the preparation of this manuscript.
THE AMERICAN
JOURNAL
OF CARDIOLOGY
Volume 55
1069
References 1. Morton WE. Occupational habituation to aliphatic nitrates and the withdrawal hazards of coronary disease and hypertension. J Occup Med 1977;19: 197-200. 2. Ramkissoon RA. Macroscopic identification of early myocardial infarction by dehydrogenase alterations. J Clin Pathol 1970;23:203-209. 3. Franciosa JA, Nordstrom LA, Cohn JN. Nitrate therapy for congestive heart failure. JAMA 1978;240:443-446. 4. Muller JE, Gunther BJ. Nifedipine therapy for Prinzmetal’s angina. Circulation 1976;57:137-139, 5. Bogaert MG. Tolerance towards glyceryl trinitrate (trinitrin) in rabbits. Arch Int Pharmacodyn Ther 1968;172:228-230. 6. Needleman P, Johnson EM Jr. Mechanism of tolerance development to organic nitrates. J Pharmacol Exp Ther 1973;184:709-715. 7. Rlnzler SH, Travel1 J, Karp D, Charleson D. Detection of coronary atherosclerosis in the living rabbit by the ergonovine stress test. Am J Physiol 1956;184:605-612. 6. Curry RC, Pepine CJ, Sabom MB, Conli R. Similarities of ergonovineinduced and spontaneous attacks of variant angina. Circulation 197967: 307-3 12. 9. McGregor M, Crelinsten GL. Possible interactions of indomethacin and nitrates in angina pectoris. Am J Cardiol 1984;53:225-227. 10. Friedman PL, Brown EJ Jr, Gunther S, Alexander RW, Barry WH, Mudge GH Jr, Grossman W. Coronary vasoconstrictor effect of indomethacin in $y5nts with coronary artery disease. N Engl J Med 1981;305:117111. Conslantlnldes R. Experimental atherosclerosis in the rabbit. In: Roberts JC Jr, Straus R, eds. Comparative Atherosclerosis. New York: Harper 8 Row. 1965276-290. 12. Maser A, Severi S, DeNes M, L’Abbate A, Chlerchta S, Marrilli M, Baltsstra AM, Parodi 0, Biagini A, Dlstante A. “Variant” angina: one aspect of a continuous spectrum of vasospastic myocardial ischemia. Am J Cardiol 1978;42:1019-1035. 13. DeServl S, Specchla G, AnQOli L. Coronary arterv spasm of different dagrees as cause of angina at rest with ST-segment depression and elevation. Br Heart J 1979:42:110-112. 14. Kerln NZ, RubentIre M, Nainl M! Wajsxcruk WJ, Pamatmat A, Cascade PN. Arrhythmias in variant angtna pectoris. Circulation 1979;60:13431350. 15. Srlachck J, Waters DD, Miller D, Therwx P. Ventricular arrhythmias during ergonovine-induced episodes of variant angina. Am Heart J 1984;107: 20-24.
Coronary Artery Spasm After Abrupt Withdrawal of Nitroglycerin in Rabbits WILLIAM C. REEVES, MD, LAWRENCE COOK, MD, MARY ANN WOOD, BS, and LAWRENCE WHITESELL, BS
The potential for abrupt withdrawal of nitroglycerin (NTG) to create coronary artery spasm was assessed in New Zealand white rabbits. In the control setting, electrocardiograms were taken from 7 anesthetized rabbits. The administration of intravenous ergonovine did not provoke ST-segment shifts or arrhythmias. Two inches of topical NTG (2 % ) was applied 3 times daily to a shaved area on the back of each rabbit over a 6-week period. Forty hours afler abrupt withdrawal of NTG, intravenous ergonovtne and indomethacin were given. Six of 7 rabbits had electrocardiographic changes: ventricular tachycardia in 2, ventricular fibrillation in 1, and significant (1 mm or more) ST-segment shifts in 5 rabbits. Three rabbits died. Sixty-four hours after NTG administration the remaining 4 rabbits were reexamined. One had baseline electrocardiographic evidence of severe myocardial ischemia. Repeat
ergonovine and indomethacin testing in the others revealed ventricular tachycardia progressing to asystole in 1, premature ventricular complexes in 1, and ST-segment elevation in another. Two of the remaining 4 rabbits died. Eighty-two hours after NTG administration the remaining 2 rabbits were found dead in their cages. Nitroblue tetrarolium studies revealed extensive myocardial infarction in both animals. Additional studies were performed in 10 normal rabbits. Neither ergonovine nor indomethacin induced ST-segment shifts or arrhythmias in this control population. It is concluded that abrupt withdrawal of NTG and subsequent administration of ergonovine and indomethacin created myocardial tschemia, ventricular arrhythmias and myocardial infarctions in these rabbits. The cause of these abnormalities is probably coronary artery spasm. (Am J Cardtol 1965;55:1066-1069)
Occupational habituation to nitrates, including withdrawal hazards of angina pectoris, coronary artery spasm and sudden death, are well d0cumented.l However, little is known regarding the safety of withdrawal of long-acting nitroglycerin (NTG) in patients treated for angina pectoris or congestive heart failure. We developed an experimental rabbit model characterized by abrupt withdrawal from long-acting NTG. This report describes our experience with this model using parenteral coronary vasoconstrictors to induce latent coronary artery spasm during the withdrawal period.
Series 1: Documentation of nitroglycerin absorption: Seven rabbits were anesthetized with intravenous sodium pentobarbital (30 mg/kg). Arterial blood pressure was measured through the central ear artery. Intravenous NTG was administered through a marginal ear vein in order to lower the mean arterial blood pressure 20 mm Hg from its baseline value. An initial dose of 10 pg/kg/min was administered over 3 minutes. This dose was doubled every 3 minutes until the appropriate blood pressure decrease was noted. Two inches of NTG ointment (2%, Kremers-Urban Company) was applied to a shaved area (16 square inches) in the interscapular region of the back 3 times daily (at 0900, 1300 and 1700 hours). During the seventh day of this treatment the rabbits were reanesthetized and the intravenous NTG protocol was repeated to the same blood pressure endpoint. Electrocardiographic analysis: Before topical NTG application, the 7 New Zealand white rabbits were anesthetized with intravenous sodium pentobarbital(30 mg/kg) and placed supine on a surgical table. Electrocardiographic (ECG) leads were attached to each extremity and the midsternum with subcutaneous needles. Simultaneous recordings of lead II and a single V lead were obtained with a Hewlett Packard Electrograph recorded at 50 mm/s using regular standardization (10 mm/mV). A baseline ECG recording was obtained. After
Methods Male New Zealand white rabbits that weighed 2.5 to 4.0 kg were used in this investigation. From the Milton S. Hershey Medical Center, Pennsylvania State University, Department of Cardiology, Hershey, Pennsylvania. Manuscript received October 23, 1984; revised manuscript received December 26, 1984, accepted December 31.1984. Address for reprints: William C. Reeves, MD, Mount Sinai Medical Center, Cardiovascular Disease Section, 950 North 12th Street, Milwaukee, Wisconsin 53233.
April 1,1985
this procedure, intravenous ergonovine (0.15 mg/kg) was given over 10 seconds. Electrocardiographic tracings were obtained immediately after ergonovine administration and at 30 seconds, 1,2,3,4 and 5 minutes after administration of ergonovine. A second intravenous ergonovine dose (0.15 mg/kg) was administered and the electrocardiogram recorded as previously described. Two inches of topical NTG (2%) was applied to the shaved interscapular area of each rabbit 3 times daily. This was done for 42 consecutive days. The interscapular area was shaved once a week during this period. Nitroglycerin administration was abruptly discontinued and 40 hours after the last administration of NTG the rabbits were reanesthetized. They were again attached to the electrograph and an initial dose of ergonovine (0.15 mg/kg) was given with ECG monitoring as previously described. This was followed by indomethacin (25 mg/kg) given intravenously over 30 seconds. Electrocardiographic monitoring was performed as with ergonovine administration. The remaining rabbits were allowed to recover and 24 hours later they were reanesthetized and ergonovine and indomethacin were again administered and monitored electrographically as at 40 hours. The remaining rabbits were returned to their cagesto recover. Myocardial infarction detection: The hearts from 2 rabbits that were found dead in their cages 82 hours after the last NTG administration were stained with nitroblue tetrazolium.2 Three transverse slices of the left ventricle were obtained from each heart. These were incubated in nitroblue tetrazolium solution for 30 minutes at 37°C. Nitroblue tetrazolium solution was made by dissolving nitroblue tetrazolium powder in distilled water at a concentration of 5 mg/ml. Ten milliliters of this solution were added to 10 ml of Sorensen’s phosphate buffer (1 M), pH 7.4 and 80 ml of distilled water. Normal heart muscle is stained by this technique. Infarcted myocardium remains unstained. Series 2: A second baseline control study was performed on 10 male New Zealand white rabbits. The rabbits were anesthetized and attached to the electrograph as previously described. Three intravenous injections were used in the following order: ergonovine (0.15 mg/kg), indomethacin (25 mg/kg), and ergonovine (0.15 mg/kg). A 5-minute ECG monitoring period was observed after each drug administration as described in series 1. Statistical analysis: The amount of intravenous NTG required to drop systemic arterial blood pressure before and after topical NTG administration was compared by the Student t test. A p value <0.05 was considered statistically significant.
Results Series 1: Nitroglycerin absorption: Initially, an average of 51 f 20 pg/kg (mean f standard error of the mean) of intravenous NTG was required to decrease mean systemic arterial blood pressure 20 mm Hg. During the seventh day of chronic NTG administration
FIGURE 2. Rabbit 5. Electrocardiogram after administration of indomethacin at 40 hours after the last dose of nitroglycerin. Upper tracing, lead II shows 2 mm of ST-segment elevation; lower tracing, lead V shows 4 mm of ST-segment elevation.
THE AMERICAN JOURNAL OF CARDIOLOGY Volume 55
1067
FIGURE 1. Rabbit 1. Electrocardiogram after administration of indomethacin at 40 hours after the last dose of nitroglycerin. Upper tracing, lead II shows downsloping ST-segment depression (2 mm) and T-wave inversion: lower tredng, a simultaneous lead V shows T-wave inversion without significant ST depression.
a significantly greater (p <0.05) amount of intravenous NTG, 95 f 30 pg/kg, was required to achieve the same decrease in systemic arterial blood pressure. This documents NTG absorption and the development of tolerance. Electrocardiographic ana2ysis: Before topical NTG administration, no significant ST-segment shifts (1 mm or greater) were noted in any rabbit at the baseline tracing or after administration of either ergonovine dose. No arrhythmias were noted. Forty hours after the last nitroglycerin dose: Ergonovine induced ST-segment elevation in 2 rabbits. Indomethacin administration resulted in ST-segment depression in 1 rabbit (Fig. l), ST-segment elevation in 4 rabbits (Fig. 2) and ventricular tachycardia and fibrillation in 2 rabbits. One rabbit had no ECG changes with ergonovine and after indomethacin administration had an immediate respiratory arrest and death with no antecedent ECG changes. Sixty-four ministration:
hours after the last nitroglycerin
ad-
Ergonovine induced isolated premature ventricular complexes in 1 rabbit and ST-segment elevation in another. Indomethacin provoked ventricular tachycardia that progressed to ventricular asystole and death in 1 rabbit. In 1 rabbit no drugs were administered because the baseline electrocardiogram exhibited 4 mm of ST-segment elevation with a reduced QRS complex in lead V and upsloping ST-segment depression in lead II (Fig. 3). Death followed shortly thereafter. Myocardial infarction detection: Transverse slices of the left ventricle from each rabbit found dead in its
1066
CORONARY
ARTERY SPASM AFTER NITROGLYCERIN WITHDRAWAL
FIGURE 3. Rabbit 5. Baseline electrocardiogram without drug provocation at 64 hours after t@ last nitroglycerin dose. Upper tracing, lead II Shows upsloping ST depression; lower tracing, lead V shows 4 mm of ST-segment elevation.
cage 82 hours after the last NTG administration had extensive myocardial infarctions (Fig. 4). Series 2: No ECG ST-segment changes or arrhythmias were observed after either ergonovine or indomethacin administrations. Discussion Sudden withdrawal after intensive occupational exposure to NTG can produce coronary artery spasm, myocardial infarction and sudden death.’ Development of similar complications caused by sudden withdrawal of long-acting NTG in patients with angina pectoris or congestive heart failure has not been well documented in published reports. Franciosa et als described 2 patients with coronary artery disease who died suddenly after discontinuing oral isosorbide dinitrate. Muller and Gunthe+ described a case of acute myocardial infarction occurring soon after withdrawal of NTG in a patient with variant angina. In clinical practice it would be difficult to establish such a causal relation between these complications and NTG withdrawal. A clinical study to evaluate the role of NTG withdrawal would be unethical. Furthermore, patients with coronary artery disease have coronary artery spasm and acute myocardial infarction and die suddenly independent of the use and withdrawal of NTG. The present study is an attempt to evaluate the
FIGURE 4. Transverse slice of the heart from a rabbit found dead in his cage 62 hours after administration of the last nitroglycerin dose. The nltroblue tetrazolium stain of the left ventricle at the papillary muscle level exhibits myocardlal Infarction (light areas), particularly evident in the ventricular septum.
role of NTG withdrawal in the development of such complications in an animal model. Caution must be used in attempting to directly extrapolate the results of our study in anesthetized rabbits to conscious humans. We selected rabbits for this investigation because they have been used successfully to demonstrate tolerance to NTG,5p6 a condition noted to occur in human vascular smooth muscle. Additionally, Rinzler et al7 developed a protocol using intravenous ergonovine and ECG monitoring to detect latent coronary artery disease in anesthetized rabbits. Ergonovine and indomethacin were chosen to test for latent coronary artery spasm in our experimental model. Ergonovine is a potent smooth-muscle vasoconstrictor commonly used to detect coronary artery spasm in human clinical studies.8 Indomethacin is a nonsteroidal anti-’ inflammatory agent. Its role in diminishing coronary flow is controversial9 However, it has been suggested to be a coronary artery vasoconstrictor, presumably through blockade of prostaglandin synthesis.1° Significant electrocardiographic ST-segment changes and ventricular arrhythmias after ergonovine and indomethacin were found only during the phase of NTG withdrawal. Six of 7 rabbits showed significant changes at the first testing interval 40 hours after the last dose of topical NTG. One rabbit had significant ST elevation after ergonovine, and 3 rabbits after indomethacin. One rabbit had significant ST depression after a 30-second episode of ventricular tachycardia after indomethacinj Another had ventricular tachycardia and ventricular fibrillation after indomethacin. At 64 hours after the last NTG administration, ST-segment shifts and ventricular arrhythmias could still be provoked with ergonovine and indomethacin. In addition, 1 rabbit had ECG evidence of severe myocardial ischemia during a baseline tracing. Only 2 rabbits survived this testing period and both were found dead in their cages the next morning. Nitroblue tetrazolium, a technique that allows macroscopic identification of early myocardial infarction by dehydrogenase alterations,2 revealed myocardial infarction in each instance. It is possible that platelet aggregation or a direct drug effect caused or contributed to the observed ahnormalities. Furthermore, we did not document histologically the absence of coronary atherosclerosis in our rabbits. However, coronary atherosclerosis probably did not play a role in the development of myocardial ischemia and infarction in our study, because rabbits are not known to spontaneously have this condition.ll We used large doses of ergonovine and indomethacin in this study. However, we do not believe that these doses created false-positive results, because neither agent provoked significant ECG abnormalities in the
April I,1985
control settings. Although our ergonovine dose (0.15 mg/kg) was larger than the usual dose administered in Rinzler’s7 series (0.05 mg/kg), it was well below the maximal dose used by him (1.0 mg/kg). The large indomethacin dose was chosen to ensure a pharmacologic effect in this animal species. Although we could not directly detect the presence of coronary artery spasm in our experimental model, the inducible ST-segment shifts, ventricular tachycardia and myocardial infarctions are suggestive of coronary artery spasm. In the setting of severe spasm, transmural ischemia results in ST-segment elevation, whereas lesser degrees of ischemia may cause ST-segment depression.12J3 Malignant ventricular rhythm disturbances are also found in humans with coronary artery spasm14 and may be induced by ergonovine in the clinical setting.15 We conclude that significant ST-segment shifts, malignant ventricular rhythm disturbances and myocardial infarction can be provoked in an experimental rabbit model after the administration of ergonovine and indomethacin during the period of immediate withdrawal from chronic NTG exposure. The probable cause of these abnormalities is coronary artery spasm. Acknowledgment: We acknowledge the expert secretarial assistance of Vicki Weisenburg and the editorial review of Dr.
Carl Christensenin the preparation of this manuscript.
THE AMERICAN
JOURNAL
OF CARDIOLOGY
Volume 55
1069
References 1. Morton WE. Occupational habituation to aliphatic nitrates and the withdrawal hazards of coronary disease and hypertension. J Occup Med 1977;19: 197-200. 2. Ramkissoon RA. Macroscopic identification of early myocardial infarction by dehydrogenase alterations. J Clin Pathol 1970;23:203-209. 3. Franciosa JA, Nordstrom LA, Cohn JN. Nitrate therapy for congestive heart failure. JAMA 1978;240:443-446. 4. Muller JE, Gunther BJ. Nifedipine therapy for Prinzmetal’s angina. Circulation 1976;57:137-139, 5. Bogaert MG. Tolerance towards glyceryl trinitrate (trinitrin) in rabbits. Arch Int Pharmacodyn Ther 1968;172:228-230. 6. Needleman P, Johnson EM Jr. Mechanism of tolerance development to organic nitrates. J Pharmacol Exp Ther 1973;184:709-715. 7. Rlnzler SH, Travel1 J, Karp D, Charleson D. Detection of coronary atherosclerosis in the living rabbit by the ergonovine stress test. Am J Physiol 1956;184:605-612. 6. Curry RC, Pepine CJ, Sabom MB, Conli R. Similarities of ergonovineinduced and spontaneous attacks of variant angina. Circulation 197967: 307-3 12. 9. McGregor M, Crelinsten GL. Possible interactions of indomethacin and nitrates in angina pectoris. Am J Cardiol 1984;53:225-227. 10. Friedman PL, Brown EJ Jr, Gunther S, Alexander RW, Barry WH, Mudge GH Jr, Grossman W. Coronary vasoconstrictor effect of indomethacin in $y5nts with coronary artery disease. N Engl J Med 1981;305:117111. Conslantlnldes R. Experimental atherosclerosis in the rabbit. In: Roberts JC Jr, Straus R, eds. Comparative Atherosclerosis. New York: Harper 8 Row. 1965276-290. 12. Maser A, Severi S, DeNes M, L’Abbate A, Chlerchta S, Marrilli M, Baltsstra AM, Parodi 0, Biagini A, Dlstante A. “Variant” angina: one aspect of a continuous spectrum of vasospastic myocardial ischemia. Am J Cardiol 1978;42:1019-1035. 13. DeServl S, Specchla G, AnQOli L. Coronary arterv spasm of different dagrees as cause of angina at rest with ST-segment depression and elevation. Br Heart J 1979:42:110-112. 14. Kerln NZ, RubentIre M, Nainl M! Wajsxcruk WJ, Pamatmat A, Cascade PN. Arrhythmias in variant angtna pectoris. Circulation 1979;60:13431350. 15. Srlachck J, Waters DD, Miller D, Therwx P. Ventricular arrhythmias during ergonovine-induced episodes of variant angina. Am Heart J 1984;107: 20-24.