Circadian variation in variant angina

Thirteen hospitalized patients with variant angina were studied to assess circadian variation in disease activity. Over 48 hours, all angina attacks were noted, a continuous Holter electrocardiogram was recorded and 2 ergonovine tests were performed 12 hours apart, 1 at 4 AM and the other at 4 PM. Only 2 patients gave a clearcut history of more frequent nocturnal or early morning attacks. During the study period, 1.8 +/- 1.6 AM and 0.62 +/- 1.2 PM angina episodes per patient were reported (p less than 0.02), but a circadian pattern was apparent in only 4 patients. However, Holter analysis revealed 5.3 +/- 13.8 AM and 2.6 +/- 8.5 PM episodes of ST elevation per patient (p less than 0.05) and 8.1 +/- 13.9 AM and 3.2 +/- 8.5 PM episodes of ST elevation, ST depression or T-wave pseudonormalization (p less than 0.01). Ten of 11 patients with Holter abnormalities had more frequent AM than PM attacks (p less than 0.01). ST elevation developed during all 13 of the 4-AM and 12 of 13 of the 4-PM ergonovine tests. In 10 cases the ergonovine threshold at which the attack occurred was lower in the morning, in no case was it lower in the afternoon, and in 3 patients the morning and afternoon doses were identical (p less than 0.01). Thus, circadian variation in disease activity both for spontaneous and provoked attacks is present in most patients with variant angina, even though it is often not clinically apparent.     

Similarities of ergonovine-induced and spontaneous attacks of variant angina.

Ergonovine has been shown to provoke attacks of variant angina, but a question remains whether spontaneous and ergonovine-induced attacks of variant angina are similar. Seven patients with variant angina undergoing cardiac catheterization were studied during transient episodes of spontaneous and ergonovine-induced rest angina with ST-segment elevation. Clinical, electrocardiographic, left ventricular hemodynamic and coronary angiographic observations were made before and repeated after ergonovine (0.05-0.2 mg I.V.). The character and duration of chest pain were similar during both spontaneous and ergonovine-induced episodes. ST-segment elevation (greater than 1 mm) was present inferiorly in three patients, anteriorly in three patients, and both inferiorly and anteriorly in one patient during both episodes. Mean heart rate and systolic arterial pressure changed little, while left ventricular end-diastolic pressure increased significantly during spontaneous or ergonovine-induced attacks. We observed subtotal or total dynamic obstruction in the left anterior descending (three patients), right coronary arteries (three patients) and both arteries in one patient during both attacks. Thus, in selected patients ergonovine-induced attacks of variant angina were remarkably similar to spontaneous episodes.     

Ergonovine testing to detect spontaneous remissions of variant angina during long-term treatment with calcium antagonist drugs

A subgroup of 22 patients with variant angina who had responded well to calcium antagonist drugs were studied to determine if ergonovine testing could help assess the need for continued therapy. Before treatment all 22 patients exhibited angina with S-T elevation during ergonovine testing done in the coronary care unit according to a previously described protocol with sequential ergonovine doses of 0.0125, 0.025, 0.05, 0.1, 0.2, 0.3 and 0.4 mg administered at 5 minute intervals. After 9.4 ± 4.7 (range 1 to 24) months of treatment (nifedipine 7 patients, diltiazem 3, verapamil 8, perhexiline 3, nifedipine and diltiazem 1), all patients were free from anginal attacks. Medication was discontinued and ergonovine testing repeated 24 to 48 hours later (3 weeks for perhexiline). In 12 of the 22 patiénts, angina or S-T segment shifts did not occur during the second ergonovine test to a maximal dose of 0.4 mg. Treatment was not restarted in these patients and all 12 remain free of variant anginal attacks 4.2 ± 2.9 (range 1 to 13) months later. In seven patients angina and S-T elevation occurred during the second ergonovine test, in the same electrocardiographic leads as during the test before treatment. In three patients the ergonovine test induced angina with S-T depression in the leads where S-T elevation had occurred during the previous test. Treatment was reinstituted in these 10 patients with a positive test. No complications resulted from ergonovine testing in any patient.We conclude that in many patients with variant angina, symptoms will disappear spontaneously and the ergonovine test will revert to negative. Treatment with calcium antagonist drugs can probably be safely discontinued in some patients with variant angina; ergonovine testing appears to be helpful in identifying such patients. Longer periods of follow-up are required to confirm that symptoms do not recur.     

Failure of ketanserin, a serotonin inhibitor, to prevent spontaneous or ergonovine-induced attacks of variant angina

Six patients hospitalized with active variant angina were treated for 3 days with the serotonin antagonist ketanserin after a 3 day control period on no medication. The number of variant angina episodes per patient per day was 1.52 +/- 1.42 during the control period and 2.05 +/- 2.30 during ketanserin therapy (p = NS). Ergonovine was administered in incremental doses of 0.0125 mg to 0.4 mg in the control period, during intravenous ketanserin administration and after 3 days of oral treatment. All 6 patients developed ST elevation during all 3 ergonovine tests. The ergonovine dose at which ST elevation developed was similar in each of the 3 periods. It is concluded that ketanserin is of no value in the treatment of variant angina and that both spontaneous and ergonovine-induced coronary spasm in man are unlikely to be mediated by a serotonergic mechanism.     

Hyperventilation and ergonovine tests in Prinzmetal's variant angina pectoris in men

Hyperventilation and ergonovine tests were carried out in a group of 30 patients with variant angina to assess the sensitivity of the 2 tests and to correlate the response with spontaneous disease activity. Hyperventilation produced a positive response in 83% (25 of 30) and ergonovine in 93% (28 of 30) of the patients. After hyperventilation 22 of 25 showed ST-segment elevation, 2 ST depression and 1 T-wave pseudonormalization; after ergonovine ST-segment elevation developed in 23 patients, ST depression in 4 and T-wave pseudonormalization in 1. In all cases the electrocardiographic changes occurred in the same leads as during the spontaneous attacks. The incidence of chest pain and ventricular arrhythmias was similar during both tests; spontaneous remission of ischemia, however, was more frequent (48 vs 14%) after hyperventilation than after ergonovine. Acute ischemia developed at a mean of 218 +/- 112 seconds after the end of hyperventilation in 19 of 25 positive tests; at that time double product was not significantly different from basal values. The sensitivity of hyperventilation was similar (95 vs 100%) to ergonovine in the patients with greater than or equal to 1 daily attack, while in those with less than 1 daily attack the sensitivity of hyperventilation decreased to 55% compared to 77% with ergonovine. Thus, in variant angina the sensitivity of both tests correlates with disease activity. Hyperventilation is a safe provocative test with a sensitivity similar to ergonovine in patients with active disease; however, in patients with sporadic attacks hyperventilation has a lower sensitivity than ergonovine and therefore a limited diagnostic value.     

Application and safety of outpatient ergonovine testing in accurately detecting coronary spasm in patients with possible variant angina

We analyzed the results of 61 consecutive outpatient ergonovine provocation tests to determine the safety and efficacy of such outpatient testing for detecting coronary artery spasm (CAS). Criteria for outpatient testing included: clinical history suggestive of variant angina, noncritical coronary artery disease documented by coronary arteriography, normal exercise treadmill test, no symptomatic arrhythmias, and no history of recent myocardial infarction. All antianginal medications were tapered and stopped. Ergonovine maleate was given as a bolus at 3-minute intervals in consecutive doses of 0.05, 0.10, and 0.25 mg. A positive test was defined as chest pain accompanied by > 0.1 mV ST segment elevation on 12-lead ECG. If pain and ST-segment elevation occurred, intravenous and sublingual nitroglycerin were immediately administered for rellef of myocardial ischemia. Of the 61 patients studied, 10 had positive tests; there were no complications. Follow-up of the 51 patients with negative studies has not revealed cardiac etiology for their chest pain. We conclude that outpatient ergonovine testing is a safe and accurate diagnostic test for identifying CAS in a highly selected population of patients with possible variant angina when performed under carefully controlled conditions.     

Ergonovine-induced myocardial ischemia: no role for serotonergic receptors?

Because ergonovine appears to produce coronary contractions by a serotonergic (5-HT) mechanism, we attempted to prevent ergonovine-induced ischemia in patients with vasospastic angina by pretreatment with ketanserin, a new selective 5-HT blocker. We studied seven patients with consistently positive results of ergonovine testing (ST segment elevation in three and ST segment depression in four). Ergonovine testing was performed before and after a bolus of 10 mg of ketanserin (all patients) and infusion of 2 to 4 mg/hr for 8 hr (six patients). To assess 5-HT blockade during ketanserin infusion, the constrictor response of hand veins to 5-HT was tested before and after ketanserin. Despite evidence of 5-HT blockade in hand veins, ergonovine-induced ischemia was not prevented by ketanserin in any patient, and there was no significant change in the dose of ergonovine required to provoke ischemia. In one patient, four spontaneous episodes of ST segment elevation occurred during infusion of ketanserin. The plasma concentrations of ketanserin at the time of ergonovine testing ranged from 61 to 127 ng/ml (mean 102) and were well above those that completely inhibit canine coronary 5-HT contractions in vitro. Although human coronary arteries may differ in their responsiveness to 5-HT or ketanserin, these data suggest that ischemia from ergonovine-induced coronary vasospasm is not mediated by 5-HT receptors.     

Increased fibrinopeptide A during anginal attacks in patients with variant angina

It is not known whether coronary vasospasm is associated with coronary thrombosis. In this study, plasma levels of fibrinopeptide A during anginal attacks in 24 patients with variant angina were examined. A hyperventilation test was used to induce angina. Hyperventilation induced angina and ST segment elevation (AST: 0.32 +/- 0.14 mV, p less than 0.01) in eight patients with variant angina. Fibrinopeptide A increased from 0.75 +/- 0.27 at control to 7.8 +/- 4.4 ng/ml (p less than 0.01) during anginal attacks in these eight patients. In addition, four patients had spontaneous attacks of angina; they also had elevated levels of fibrinopeptide A during attacks (from 2.0 +/- 1.2 at control to 21.9 +/- 18.0 ng/ml [p less than 0.01] during attacks). Hyperventilation did not induce either angina or ST segment elevation in 12 of the patients with variant angina. Fibrinopeptide A levels did not change with hyperventilation in these patients. To determine whether elevated plasma levels of fibrinopeptide A were associated with angina, the plasma levels of fibrinopeptide A were examined during exercise-induced angina in seven additional patients with stable effort angina. They all developed angina with treadmill exercise; however, plasma fibrinopeptide A did not change. Therefore, only the patients with variant angina demonstrated elevated levels of fibrinopeptide A during anginal attacks. These findings suggest that coronary vasospasm associated with myocardial ischemia may induce stasis of blood, resulting in fibrinogen-fibrin conversion in the coronary vessels.     

Limited efficacy of magnesium for the treatment of variant angina.

Some patients with variant angina show both ST segment elevation at rest and exercise-induced ST segment elevation. Magnesium deficiency has also been observed in patients with variant angina. This study investigated the correlation between the degree of magnesium deficiency and the efficacy of intravenous administration of magnesium in patients with variant angina. Fifteen patients with angiographically confirmed variant angina were assessed for magnesium deficiency and whether intravenous administration of magnesium (19.2 mEq/l) suppressed exercise-induced ST segment elevation. All 15 patients were studied with a magnesium retention test (0.2 mEq/kg over 4 hr) to analyze magnesium deficiency. In our study, magnesium retention rate in patients with variant angina was not higher than that of controls (57 +/- 24% vs 45 +/- 10%, NS). All 15 patients had anginal attacks during accelerated exercise combined with hyperventilation after placebo infusion, whereas only 8 patients had anginal attacks after magnesium administration. ST segment elevation occurred in 14 patients after placebo infusion, but in only 4 patients after magnesium administration. There were no correlations between disease activity, degree of magnesium deficiency or failure of suppression of ST elevation by the intravenous administration of magnesium. Intravenous administration of magnesium can suppress exercise-induced coronary spasms in some patients with variant angina, but the degree of magnesium deficiency did not correlate with the suppressions of exercise-induced ST elevation after magnesium administration. Intravenous administration of magnesium had limited efficacy in patients with variant angina and exercise-induced ST segment elevation.     

Transient changes in left ventricular mechanics during attacks of Prinzmetal's angina: an M-mode echocardiographic study

M-mode echocardiograms were recorded in 12 patients with Prinzmetal's angina during 29 episodes of transient myocardial ischemia at rest (18 spontaneous and 11 ergonovine-induced). At peak ST segment elevation a regional mechanical impairment was observed in the interventricular septum during 23 episodes of angina and in the posterior wall during six episodes. In the 18 spontaneous episodes the left ventricular ischemic wall, when compared to the basal state, was found to have a significant reduction in motion (-76.3 +/- 9.1%) (mean +/- SEM), in diastolic thickness (-11.7 +/- 2.5%), and in percent systolic thickening (-88.0 +/- 5.6%). Increase in left ventricular end-diastolic diameter (+13.1 +/- 2.1%) and decrease in percent fractional shortening (-38.1 +/- 3.7%) were also observed. When ST segment was back to the isoelectric line, a transient overshoot in regional left ventricular function was observed. In induced episodes statistically significant changes could be detected by M-mode echocardiography even before appearance of ST segment elevation and anginal pain. No significant difference was found in type or degree of mechanical impairment between induced and spontaneous episodes. Therefore, in patients with Prinzmetal's angina: (1) M-mode echocardiography allows detection of mechanical changes due to transient myocardial ischemia; and (2) mechanical impairment occurs earlier than clinical (pain) and electrocardiographic (ST segment elevation) signs of transmural ischemia.     

Failure of transdermal nitroglycerin to improve chronic stable angina: a randomized, placebo-controlled, double-blind, double crossover trial

We assessed the effect of transdermal nitroglycerin (NTG-TTS), releasing 5 mg/24 hr, in 11 patients with chronic stable angina during a randomized, placebo-controlled, double-blind, double crossover trial of four 1-week periods. All patients had a positive exercise test and coronary artery disease prior to entry into the study. Efficacy was assessed weekly by anginal diaries, ambulatory ST segment recordings, and computerized exercise testing 2 to 4 hours after renewal of NTG-TTS. One patient withdrew in the first week (placebo) and was excluded from all analysis. The weekly frequency of anginal attacks was 9 +/- 11 (mean +/- SD) and 9 +/- 10 during the 2 placebo weeks and 11 +/- 14 and 9 +/- 11 during NTG-TTS; the time to 1 mm ST segment depression (seconds) was 268 +/- 178 and 303 +/- 217 with placebo and 228 +/- 221 and 285 +/- 178 with NTG-TTS; exercise duration (seconds) was 375 +/- 230 and 467 +/- 254 during placebo and was 394 +/- 233 and 412 +/- 236 during NTG-TTS. The weekly number of episodes of ST depression was 11 +/- 9 and 8 +/- 5 during placebo and 8 +/- 5 and 9 +/- 7 during NTG-TTS. Our study failed to show any consistent differences in patients treated with placebo or NTG-TTS.     

Coronary arteriography and left ventriculography during spontaneous and exercise-induced ST segment elevation in patients with variant angina

The present study is an angiographic demonstration of coronary artery spasm during both spontaneous and exercise-induced angina in three patients with variant angina. In each case, clinical, ECG, coronary angiographic, and left ventriculographic observations were made at rest, during spontaneous angina, and during exercise-induced angina. The character of chest pain was similar during spontaneous and exercise-induced episodes. ST segment elevation was present in the anterior ECG leads during both episodes. The left anterior descending coronary artery became partially or totally obstructed during both types of attacks. When coronary spasm was demonstrated during both types of attacks, left ventriculography disclosed akinetic or dyskinetic wall motion in the area supplied by the involved artery. In those patients with reproducible exercise-induced ST segment elevation and chest pain, thallium-201 scintigraphy showed areas of reversible anteroseptal hypoperfusion. Thus in selected patients exercise-induced attacks of angina were similar to spontaneous episodes.     

Ergonovine testing in a coronary care unit

This study describes the results of ergonovine testing in 100 consecutive patients who underwent this procedure in a coronary care unit. All patients had recently undergone coronary arteriography. A bolus injection of ergonovine was administered at 5 minute intervals in the following doses (mg): 0.0125, 0.025, 0.05, 0.1, 0.2, 0.3 and 0.4. The criterion for a positive test was the appearance of S-T elevation greater than 1 mm. The test was positive in all 17 patients known to have variant angina and in 18 (40 percent) of 45 patients who had a history of chest pain judged strongly suggestive of variant angina but who had no electrocardiogram recorded during pain. Of 38 patients with a history of chest pain classified as not entirely typical of variant angina, only 1 (2.6 percent) had a positive test.Of the 64 patients with a negative ergonovine test, 47 had chest pain and 25 had nausea but none had more serious complications. Ventricular arrhythmia accompanied S-T elevation in 18 of the 36 patients with a positive test but occurred in only 4 of the 64 with a negative test (p < 0.0005). No patient needed treatment with antiarrhythmic drugs. Four of the 36 patients with a positive test had serious complications: severe translent hypotension (2 patients), recurrent episodes of angina with S-T elevation (1 patient) and a subendocardial infarction (1 patient). Thus, ergonovine testing is useful in patients with a typical clinical history of variant angina but without an electrocardiogram recorded during pain. in this study, a small but definite incidence of serious complications occurred during a positive test.     

Vasospastic ischemic mechanism of frequent asymptomatic transient ST-T changes during continuous electrocardiographic monitoring in selected unstable angina patients

Asymptomatic episodes of ST segment and/or T wave changes are often reported during Holter monitoring in patients with angina pectoris. However, the interpretation of such changes is debated relative to silent myocardial ischemia. We studied 11 patients admitted to the CCU because of frequent episodes of unstable anginal attacks who had undergone repeated periods of Holter monitoring, characterized by predominantly occurring asymptomatic episodes of ST segment and/or T wave changes associated with less frequent typical anginal attacks. In a total of 89 days of Holter monitoring, the patients evidenced 520 episodes of transient ECG changes including 180 of ST elevation, 73 of ST depression, and 267 of T wave alterations. Only 12% of episodes were symptomatic. Coronary injection during asymptomatic ST-T changes was performed in eight patients. In six it was possible to document spontaneous coronary spasm. In seven patients ergonovine administration induced anginal pain, ST-T changes, and coronary spasm. In all patients the anginal attacks completely disappeared with medical treatment and the asymptomatic episodes were abolished in six and reduced in four. Our findings support the hypothesis that in certain selected unstable anginal patients, transient asymptomatic ECG changes are caused by acute myocardial ischemia.     

Clinical characteristics of patients with variant angina complicated by myocardial infarction or death within 1 month

Of 132 consecutive patients hospitalized during a 5 year period because of active variant angina, 18 died or had a myocardial infarction within 1 month. In 4 patients an episode of pain and S-T elevation unrelieved by calcium antagonist drugs and intravenous nitroglycerin persisted for more than 1 hour, inducing cardiogenic shock and death before the appearance of Q waves and elevated serum enzyme levels. In the other 14 patients myocardial infarction developed in the electrocardiographic leads in which S-T elevation had occurred during attacks of variant angina.Clinical features were not helpful in distinguishing the 18 patients with complications from the other 114. Angina at rest had been present for less than 1 month in 7 of the 18 patients with infarction compared with 31 of 114 in the other group (probability [p] not significant [NS]). Before infarction the artery presumed to be perfusing the involved territory contained a fixed stenosis of 70 percent or more of luminal diameter in 8 of the 14 patients with complications who had coronary arteriograms compared with 50 of 112 in the other group (p = NS). In 13 of the 18 patients, complications occurred in spite of large doses of calcium antagonist drugs. In 11 of these 13, attacks of variant angina were monitored for 3 to 17 days both before and during treatment. All 11 had fewer attacks with treatment and 5 had no attacks. Daily attacks per patient decreased from 4.6 ± 4.3 to 0.5 ± 0.7 (mean ± standard deviation) (p < 0.01). It is concluded that in variant angina of recent onset myocardial infarction occurs frequently and unpredictably. Myocardial infarction may occur in the absence of severe fixed lesions and in spite of apparent clinical improvement with administration of calcium antagonist drugs.     

Maximal exercise testing in patients with spontaneous angina pectoris associated with transiet ST segment elevation. Risks and electrocardiographic findings.

Six patients with spontaneous angina associated with transient ST segment elevation had a multistate maximal exercise (bicycle) test. In 5 patients, typical electrocardiographic changes were recorded during exercise, namely ST segment elevation often accompanied by an increase in the voltage of the R wave and a widening of the QRS complex. Four of these patients developed severe rhythm disturbances: ventricular tachycardia (2 cases) and ventricular flutter (1 case) were the reason for early interruption of the test in 3 patients, while 1 patient had a short run of ventricular tachycardia after exercise. These rhythm disturbances which spontaneously regressed in all cases were consistently preceded by obvious ST elevation and in 2 patients were attended by slight chest discomfort. Maximal exercise testing of patients suspected of variant angina provides important diagnostic information in many patients, but the risks of potentially lethal arrhythmias should be considered and resuscitation facilities should always be immediately available.     

The nature of ventricular arrhythmias during ergonovine-induced vasospastic angina pectoris

The peak incidence of ventricular fibrillation in acute myocardial infarction usually occurs during the first hours after the onset. Electrophysiological changes immediately after the onset have been studied in animal models, but are still incompletely understood in humans. For clarification of the characteristic features of ventricular arrhythmias during acute myocardial ischemia, ventricular arrhythmias were studied in 81 patients with vasospastic angina pectoris induced by ergonovine. Ventricular arrhythmias occurred in 45 of these patients, including ventricular tachycardia in 15, and ventricular fibrillation requiring repeated DC defibrillation in two patients. Most ventricular extrasystoles occurred before the ST segment reached maximum elevation, while reperfusion arrhythmias were less common. In many patients the coupling intervals varied, and the configuration was multiform. It is concluded that ventricular arrhythmias occurring during ergonovine-induced coronary spasm show different characteristics from those occurring during chronic ischemia. As the arrhythmias in this study seem, in some ways, to resemble arrhythmias occurring at the onset of myocardial infarction, the results might provide useful information on ventricular arrhythmias in myocardial ischemia in humans.     

Ventricular arrhythmias in patients with rest angina: correlation with ST segment changes and extent of coronary atherosclerosis

Major ventricular arrhythmias occurring concurrently with myocardial ischemia are presumed to be the most frequent mechanism for sudden cardiac death. Two hundred eighteen catheterized patients with angina pectoris at rest were reviewed to identify clinical, ECG, and arteriographic features that might correlate with the presence of serious ventricular arrhythmias occurring during episodes of rest pain. Ventricular arrhythmias during episodes of rest pain were significantly more common in patients who manifested transient ST segment elevation in the anterior leads and in patients with marked transient ST segment shifts (greater than 5 mm). Ventricular arrhythmias during episodes of rest pain were not more common in patients with extensive coronary artery disease.     

Intravenous nitroglycerin suppresses exercise-induced arrhythmias in patients with ischaemic heart disease: implications for long-term treatment.

We studied 20 patients with ischaemic heart disease, who consistently developed complex ventricular arrhythmias during exercise testing. Treadmill exercise was performed twice, both during the placebo infusion and then during intravenous administration of nitroglycerin, titrated to reduce systolic blood pressure by 10 mmHg. Exercise duration in those administered placebo was 7.8 +/- 1.7 and 7.9 +/- 1.5 min, respectively (ns); angina developed in five patients and ischaemic ST changes in 10. In those administered nitroglycerin, exercise duration increased to 8.4 +/- 2 min (P less than 0.05). Diagnostic ST segment depression was observed in only two patients and only one had angina. Ventricular arrhythmias, consistently present during both tests on those administered placebo, were dramatically reduced by nitroglycerin in all 20 patients. There were 455 (mean 35.8 +/- 16.8) and 418 (mean 34.4 +/- 11.1) ventricular ectopic beats in the two exercise tests on those administered placebo and 11 in those receiving the nitroglycerin infusion (mean 0.6 +/- 0.1) (P less than 0.001). There were 28 and 29 couplets in those receiving placebo (ns) and none in those receiving nitroglycerin (P less than 0.001). Ventricular tachycardia was present in six and eight patients who received placebo but in none in those administered nitroglycerin (P less than 0.001). Abolition of exercise-induced arrhythmias was maintained during chronic treatment with oral coronary vasodilators. Prevention of exercise-related arrhythmias by nitroglycerin appears a good indicator of their ischaemic origin and may provide valuable information for long-term prophylaxis with oral vasodilators, thus avoiding antiarrhythmic agents with their potential side effects.     

Characterization and therapy of ventricular arrhythmias due to transient myocardial ischemic attacks in variant angina

Forty-two patients with variant angina were studied by ambulatory ECG monitoring to determine the incidence and the characteristics of ventricular arrhythmias during ischemic attacks. Twenty-six patients had no ventricular arrhythmias in 633 ischemic attacks; 16 patients had ventricular arrhythmias in 116/586 ischemic attacks. The number of ischemic attacks per day and the magnitude of ST elevation were significantly (p less than 0.05) greater in patients with ventricular arrhythmias. Ventricular arrhythmias appeared at the onset or at the peak of ST elevation (first phase) in 17 ischemic attacks, during the resolution of ST elevation (second phase) in 43 attacks, during both the phases in 9 attacks. ST alternans appeared during 6 ischemic attacks with arrhythmias. Two episodes of ventricular fibrillation and 22 runs of ventricular tachycardia occurred during the first phase, 17 episodes of ventricular tachycardia were recorded during the second phase. Ventricular tachycardia of the second phase compared with ventricular tachycardias of the first phase were significantly (p less than 0.01) slower, uniform and initiated by a late premature beat. Incidence of arrhythmias of the second phase was strictly correlated with the duration of ischemic attacks. Nine patients who showed ventricular arrhythmias during the second phase of ischemic attacks were enrolled in a cross-over study to assess the antiarrhythmic effects of nifedipine (120 mg/day) and verapamil (480 mg/day). During treatment with nifedipine, the frequency of ischemic attacks declined by 85%, while the frequency of attacks with arrhythmias declined by 97% (p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)