Role of Nitrates
in Acute Myoqxudial John T. Flaherty,
In patll with acute myocardlal infarction, intravenous nitroglycerin lowers left ventricular filling pressure and systemic vascular resistance. At lower infuslon rates (46 pg/min) nitroglycerin is principally a venodilator, whereas at hiier infusion rates more balanced venous and arterial dilating effects are seen. Patients with left ventricular failure demonstrate increased or maintained stroke volumes, whereas patients without failure show a decrease in stroke volume. All hemodynamic subgroups show a decrease in left ventricular filling pressures and a reduction in electrocardiographic evidence of regional myocardiil ischemia. Longer-term infusions (20-48 hours) have been shown to result in myocardial preservation, as assessed by global and regional left ventricular functlln and laboratory indices of infarct s&e. Comparison of intravenous nitroglyercin and sodium nttroprusskle reveals increased intercoronary collateral flow with nitroglyercin, in contrast to a decrease with nitroprusskle, compatible with a “coronary steal.” Short-term administration of intravenous nttroglyercin with or without chronic administration of long-acting &rates have been found both to reduce short-term mortality and to have longterm beneficial effects on left ventricular remodeling in patients with anterior transmural infarctions. Current clinical practice woukl utilize intravenous nttroglyercin as inttiil therapy for patients receivhrg intravenous thrombolytic therapy and/or acute percutaneous transluminal coronary angioplasty within 4-6 hours of the onset of symptoms of acute myocardial infarction, in order to optimize intercoronary collateral flow until reperhrsion can be accompliihed. Patients reachhtg the hospital >6 hours but c 14 hours From the Division of Cardiology, Johns Hopkins of Medicine, Baltimore, Maryland. This article from Chapter 12, “The Role of Nitrates in Infarction and Post-Infarction,” by J.T. Flaherty, Pepine CJ, Thadani U, eds. Medical Therapy Disease: Nitrates, Beta Blockers, and Calcium ton: Little, Brown, 1992. Address for reprints: John T. Flaherty, Cardiology, Johns Hopkins University School more, Maryland 2120.5.
University School has been modified Acute Myocardial MD, in Abrams J, of Ischemic Heart Antagonists, BosMD, Division of of Medicine, Balti-
Infarction
MD
after symptom onset can still benefit from intravenous nttroglyercin for 2446 hours. Patients developi~ congestive heart failure and/or severe arterial hypertension as compliiins of ad acute i~rctlon as well as patients developing postinfarction unstable angina are also candidates for Intravenous nttroglyercin therapy. As wlth the a-ensin-converting enzyme inhibitors, r6 weeks of posthtfarction treatment with long-acting nltrates woukl also appear to reduce left ventricular remodeling in patients with anterior transmural infarctions. (Am J Cardiol166~70:726-6lB)
A
s recently as 1966 Friedberg’s Textbook of Cardiology warned against the adminiskation of nitroglycerin to patients with acute myocardial infarction. l Despite these warnings, cardiologists at the Johns Hopkins Hospital began studying the short-term hemodynamic and antiischemic effects of intravenous nitroglycerin in patients with acute myocardial infarction > 1.5 years ago. 2-5 More recently, we performed a randomized, placebo-controlled trial of intravenous nitroglycerin administered for 48 hours with thie goals of reducing infarct size and improving clinical outcomes.6 On October 1, 1981, intravenous nitroglycerin was released for general clinical use by the U.S. Food and Drug Administration. The clinical indications for which the intravenous preparation was approved include unstable angina and left ventricular failure complicating acute myocardial infarction. This article will review: (1) the acute hemodynamic responses to short-term infusion in patients with acute myocardial infarction; (2) the results of our placebo-controlled trial of long-term (48-hour) infusion of intravenous nitroglycerin in patients with acute myocardial infarction; (3) the results of other published clinical trials utilizing intraven0u.s nitroglycerin in an attempt to reduce infarct size; (4) a comparison of the hemodynamic and intercoronary collateral flow effects of nitroglycerin and nitroprusside; (5) the results of 2 placebo-con,A SYMPOSIUM:
NITROGLYCERIN
THERAPY
731B
trolled clinical trials of nitroprusside in acute myocardial infarction; and (6) the results of our study of the effects of acute administration of intravenous nitroglycerin in combination with intraaortic balloon counterpulsation therapy in patients with large transmural infarctions. Finally, I will present my own current recommendations for the use of intravenous nitroglycerin in patients with acute myocardial infarction. ACUTE HEMODYNAMIC EFFECTS OF SHORT-TERM INFUSION IN PAllENTS WITH ACUTE MYOCARDUU. INFARC’RON In 1972 cardiologists at the Johns Hopkins Hospital began administering intravenous mtroglycerin to patients with acute myocardial infarction with the goal of reducing regional myocardial ischemia and, when possible, improving left ventricular hemodynamics. In order to be sure that by lowering coronary perfusion pressure, regional myocardial ischemia was not being made worse, precordial ST-segment monitoring was employed. In the first 12 patients that we studied, only a small mean blood pressure decrease was induced (average -7 mm Hg), while at the same time a significant lowering of left ventricular filling pressure was documented.2 In all 12 cases a reduction in the sum of the precordial ST-segment voltages was observed.
Left Ventricular Filling (mm Hg)
Pressure
RGURE I Hemodynamk effects of intravenous nltroglycerln in the presence and absence of left ventdcular (LV) failure. Stroke volume Index (ml./mZ) Is plotted versus LV flllhlg pressure. llle responses obtained WlRl lntvavetlous Mroglycerln In bullvlduol hemodynamk subgroups are hdkated by arrows. A family of hypothetlcol Starllllg ventrkular timcUon cuwee Is IBed by doshed curves wlth dlmlnished contractllRy expressed by d ownward and rlghtward dlsplacemont. The vertical do+hed line ln&ates an upper llmlt of normal LV fllllng pressure. 746
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
70
Reassured by this initial experience, higher infusion rates of nitroglycerin were used in another clinical trial, which resulted in greater (15-30 mm Hg) reductions in mean arterial pressure.3 Beneficial anti-ischemic effects were observed in all 30 patients, irrespective of the presence or absence of left ventricular failure, as assessed by precordial ST-segment mapping. By stepwise titration of the intravenous infusion, nitroglycerin was shown to act principally as a venodilator at lower infusion rates, lowering left ventricular filling pressure by a mean of 10 mm Hg (45%), while causing minimal lowering of mean arterial pressure. At higher infusion rates, more balanced venous and arterial dilating effects were observed. In association with a similar lowering of left ventricular filling pressure (52%), a 20% lowering of mean arterial pressure was observed. Effects on left ventricular hemodynamics varied with the degree of underlying left ventricular failure (Figure 1). Patients without hemodynamic evidence of left ventricular failure demonstrated both a decrease in stroke volume and a decrease in the left ventricular filling pressure, evidencing the predominant preload lowering or “diuretic-like effect” induced by venodilation. In contrast, patients with mild left ventricular failure, as evidenced by an elevated left ventricular filling pressure but a normal stroke volume, demonstrated a lowering of left ventricular filling pressure but with a maintenance of stroke volume, suggesting that nitroglycerin was inducing a degree of afterload lowering as well. Patients with hemodynamic evidence of a more severe degree of left ventricular failure, as evidenced by both an elevated filling pressure and a reduced stroke volume, demonstrated the most beneficial hemodynamic effects. In this subgroup of patients, left ventricular filling pressure was again lowered, but in addition stroke volume was increased, compatible with a more balanced arterial and venodilating eflFect (Figure 1). Similar differential effects on hemodynamics have been reported for nitroprusside. This differential effect of nitroglycerin on stroke volume is most likely the result of differences in arterial impedance in patients with and without heart failure (Figure 2). Patients with moderate or severe left ventricular failure have been shown to have increased sympathetic tone, increased circulating catecholamines, and elevated systemic vascular resistance; the arterial-arteriolar actions of nitroglycerin are particularly useful in subjects with impaired left ventricular contractile function. SEPTEMBER
24, 1992
In summary, short-term infusion of intravenous nitroglycerin to patients with acute myocardial infarction resulted in improvement in electrocardiographic evidence of regional ischemia in all hemodynamic subgroups. Patients with the most severe degree of left ventricular failure demonstrated the most beneficial hemodynamic effects, obtaining both a reduction in pulmonary venous congestion and an improvement in forward cardiac output. Careful upward titration of the infusion rate allowed these beneficial anti-ischemic and hemodynamic effects to be obtained without excessive lowering of coronary pressure and without inducing a reflex tachycardia. In fact, the mean heart rate during intravenous nitroglycerin infusion was lower than control in all hemodynamic subgroups, most notably in the subgroups of patients with evidence of left ventricular failure. We also performed 2 additional acute hemodynamic studies in an effort to determine whether or not the anti-ischemic effects of intravenous nitroglycerin could be augmented by the addition of phenylephrine, which by maintaining coronary perfusion pressure might augment the anti-ischemic effects of nitroglycerin alone, or of propranolol, which by adding negative inotropic and negative chronotropic effects might augment the antiischemic effects of nitroglycerin by further reducing myocardial oxygen demand. The addition of the ol-adrenergic agonist and vasoconstrictor, phenylephrine, not only reversed the afterload-lowering effects, but also increased left ventricular filling pressure and reversed the beneficial anti-ischemic effects obtained with nitroglycerin alone.4 Addition of propranolol also resulted in a reversal of the preload-lowering effect obtained with nitroglycerin alone.5 Further, when the short-term nitroglycerin infusion was discontinued, left ventricular filling pressures rose above control levels, especially in patients with significant left ventricular dysfunction, reflecting the very short half-life of parenteral nitroglycerin (3 minutes) compared with the 6-hour elimination half-life of parenteral propranolol. RESULTS OF A RANdoMlZED PLACEBO-CONTROLLED CUNICAL TRW. OF LONG-TERM INFUSION OF NlTROGLYCERlN IN PATIENTS WITH ACm MYOCARDW INFARCTION
Based on the positive hemodynamic and antiischemic effects obtained on our initial short-term infusion studies,2,3 a randomized prospective trial of long-term infusion of nitroglycerin in patients with acute myocardial infarction was undertaken.6
Patients received a 48-hour infusion of eithe.r intravenous nitroglycerin or placebo, followed by the application of nitroglycerin or a placebo ointment for an additional 72 hours. Included in thi.s study were 104 patients with chest pain and electrocardiographic changes suggestive of acute myocardial infarction who presented < 12 hours after the onset of symptoms. Prior to initiation of the study drug infusion, we obtained a thallium-201 myocardial perfusion scan and a technetium-99m gated blood-pool scan. A computer-assisted scoring technique was used tie define a thallium defect score that provided a quantitative index of the extent of myocardium ait risk. It was recognized that the areas of decreaseld thallium uptake included both presently and previously infarcted myocardium as well as reversibly ischemic myocardium. The pretreatment gateld blood-pool scan (or, in a few patients, a 2-dimensional echocardiogram) defined baseline left ventricular function. The nitroglycerin infusion was begun at 5-10 pg/min, and the infusion rate increased stepwise until a 10% lowering of mean arterial pressure wa.s obtained. Blood pressure was monitored noninvalsively using an ultrasonic blood pressure cuff. Once this hemodynamic endpoint had been reached, the
I I
j
I
I
I I I I
I
, I I I
\ Severe
LV Dysfunction
INCREASING SYSTEMIC VASCULAR RESISTANCE
-
FIGURE 2. Dlfferentlal effects of afterload reduction on stroke volume according to hemcdynamlc subg~~p. Curves shown are for patients wRh normal left ventricular (LV) functkn as well as for patients wlth mlkl, moderate, and severe LV failure. Vertkal dashed llnes lndkzate pretmatmeht (right) and post-treatment (left) levels ofsystemk vascular resistance (SVR). Vertkal arrows lndkate the msignltude of the Increase In stroke volume lmlcrx that would be expected from this reduction In SVR for each hemodynamlc subgroup. A SYMPOSIUM:
NITROGLYCERIN
THERAPY
751B
final infusion rate was continued for 48 hours, unless side effects or hypotension required reduction or discontinuation of the infusion. The mean infusion rate required to reach the 10% mean blood pressure lowering endpoint was 90 pg/min. After 48 hours, the nitroglycerin infusion was discontinued, and 1 inch of nitroglycerin ointment was applied every 4 hours, with the dose increased stepwise, in an attempt to maintain the level of arterial pressure previously obtained with intravenous therapy. Control patients had 1 inch of placebo ointment applied every 4 hours. The thallium-201 gated blood-pool scans (or 2-dimensional echocardiogram) were obtained 7-14 days later. Of the 104 patients, 56 were randomized to intravenous nitroglycerin and 48 to placebo. There were no significant differences in admission clinical or scintigraphic parameters between the groups. Patients were divided into early- and late-treatment subgroups, according to the time interval between the onset of symptoms of acute infarction and time of the initiation of treatment. Patients treated < 10 hours from the onset of chest pain had “early” treatment, and those treated 210 hours after symptom onset had “late” treatment. The retrospective time cutoff of 10 hours divided the nitroglycerin-treated patients into 2 equal subgroups. Three-month mortality tended to be lower in the early nitroglycerin treatment subgroup (14%) compared with mortality in the other 3 treatment subgroups (21-28%). However, perhaps because of the small study population, this difference approached but did not reach statistical significance. When the unfavorable clinical outcomes of inhospital death, infarct extension, or development of new congestive heart failure were examined separately, the incidence of each tended to be lower with early nitroglycerin treatment, but again
PLACEBO 12 mm
10
47
48
LVFP
J CK infarct
SAP 10%
1 CK infarct size 37% but for inferior transmural infarcts only
SAP 20 mm Hg
J in-hospital mortality from 23% to 5% and precordial R waves better preserved in anterior infarcts
MAP 10%
Reduced infarct expansion, CK infarct size and & 3 month mortality
Inclusion Criteria
Patients (n) 104
60
Duration (hr)
85
Hg All infarcts
6
54
24
24
All infarcts
10
51
24-168
5
45
310
Only transmural
CK = creatine kinase; EF = ejection fraction; pressure; & = decrease.
IV = intravenous:
infarcts
LVFP = left ventricular
artery diastolic or pulmonary capillary wedge pressures of 2 12 mm Hg, transmural infarction, and an increase in total creatine kinase (CK) levels. Treatment was instituted a mean of 10 hours after the onset of symptoms, and nitroglycerin was titrated to lower the elevated pulmonary artery diastolic pressure; excessive lowering of arterial pressure was avoided. Total CK infarct size was found to be 23% lower for nitroglycerin-treated versus control patients (p 1.5 mm and a persistent R wave. A recent study by Jugdutt and WarnicalO reported beneficial effects of acute administration of intravenous nitroglycerin on early mortality as well as left ventricular ejection fraction, extent of left ventricular asynergy, and derived indices of infarct expansion and thinning. These data are discussed in Jugdutt’s review in this symposium. These investigators randomized 310 patients to intravenous nitroglycerin infusion or placebo for an average of 39 hours.ll Serial 2-dimensional echocardiography was employed to assess the incidence and severity of infarct expansion. In addition to demonstrating a reduction in expansion of the infarct segment when reassessed 10 days after admission, this large placebo-controlled trial was able to demonstrate significant reductions in CK infarct size and 3-month mortality, as well. The reduction in death was seen exclusively in the subjects with anterior wall infarction. Based on the premise that the majority of th.e published controlled clinical trials of intravenous A SYMPOSIUM:
NITROGLYCERIN
THERAPY
778
TABLE II Nitroprusside
Comparison in Acute
of Intravenous Myocardial
Nitroglycerin Infarction
with
Nitroglycerin Chairello
Nitroprusside
et alI3
Patients with anterior transmural myocardial infarction
j ST elevations
Open chest anesthetized dog model with acute coronary ligation
& ST elevations f MBF
Mann et alI4 Patients with angiographically visible intercoronary channels MBF = myocardial
t ST elevations
and
t MBF
t ST elevations 1 MBF
and
1 MBF
blood flow; 1 = decrease;
t = increase.
nitroglycerin in acute myocardial infarction failed to detect an unequivocal reduction in mortality due to their relatively small sample sizes, Yusuf and Collinsl* pooled mortality data from 7 randomized placebo-controlled studies. Although only 2 of the 7 trials9J1 were able to demonstrate a statistically significant reduction in mortality, pooling the mortality data from all 7 studies, which included > 1,000 patients, revealed a highly significant 30% reduction in mortality in the intravenous nitroglycerin-treated patients versus control patients. COMPARISON OF THE EFFECTS OF NITROGLYCERIN AND NlTROPRUSSlDE ON REGIONAL ISCHEMlA AND INTERCORONARY COLLATERAL BLOOD FLOW
Several experimental animal and clinical studies have compared the hemodynamic and anti-ischemit effects of nitroglycerin and nitroprusside (Table II). These suggest that nitroprusside may induce a coronary steal phenomenon with subsequently differing effects on the myocardium, compared with beneficial anti-ischemic actions of nitroglycerin.13-l6 In order to define the mechanisms responsible for the apparent opposite effects of nitroglycerin and nitroprusside on regional ischemia, Chiarello et all3 carried out a study in an open chest canine model. Following ligation of the left anterior descending coronary artery, nitroprusside or nitroglycerin were titrated to lower mean arterial pressure by 200 mm Hg. Utilizing the radioactive microsphere technique, myocardial blood flow distal to the ligated left anterior descending coronary artery was found to decrease and epicardial ST-segment elevations to increase during nitroprusside infusion. In contrast, during nitroglycerin infusion, ischemic zone myocardial blood flow increased and 78B
THE
AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
70
epicardial ST segment elevations decreased. Since the coronary artery remained ligated, the improvement in regional ischemia observed with nitroglycerin must be taken as evidence of an increase in intercoronary collateral blood flow. In contrast, during nitroprusside infusion, collateral flow to the ischemic zone actually decreased, and flow to the nonischemic zone increased, providing evidence of a coronary steal. In the cardiac catheterization laboratory, Mann et all4 used the xenon-133 washout technique to study the effects of nitroglycerin and nitroprusside on regional myocardial blood flow. In a subgroup of patients with fixed coronary artery disease and angiographically visible intercoronary collaterals, myocardial blood flow distal to a severe coronary artery stenosis was found to increase with nitroglycerin and to decrease with nitroprusside. Cappuro et a1,15using an open chest anesthetized dog model with well-developed collateral vessels induced by previous implantation of an amaroid constrictor, demonstrated greater sensitivity of intercoronary collateral channels to the vasodilating effects of nitroglycerin than of nitroprusside. Finally, Marcho and Vatner16 demonstrated in a conscious, instrumented dog model that, compared with nitroprusside, nitroglycerin induces relatively less dilation of the small resistance vessels and greater dilation of the large (conductance) coronary vessels, which include intercoronary collateral channels. CUNICAL TRlALS OF NCTROPRUSSIDE ACUTE MYOCARDW INFARCTlON
IN
In 2 published prospective randomized, placebocontrolled trials, nitroprusside or a placebo was administered to patients with acute myocardial infarction in an attempt to reduce infarct size and/or improve short-term mortality (Table III). In a large multicenter Veterans Administration Cooperative Trial, nitroprusside was titrated to reach 1 of 4 endpoints: (1) a reduction of left ventricular filling pressure to < 60% of control; (2) a lowering of systolic arterial pressure to 20% of the control pressure plus 76 mm Hg; (3) the development of significant side effects; or (4) a maximum infusion rate of 200 pg/min.18 Included were 812 patients with the onset of chest pain < 24 hours prior to admission, and transmural infarction, as evidenced by ST-segment elevations and subsequent development of pathologic Q waves or a new conduction defect (Table III). Approximately 67% of patients had anterior wall infarctions. A left ventricular filling pressure of r 12 mm SEPTEMBER
24,
1992
TABLE
III Clinical
Trials of Intravenous
Nitroprusside
in Acute Myocardial
Clinical Trial
Infarction Late Treatment Group
Early Treatment Group
Cohn et alI8 812 patients, mean time of Rx 17 hours after onset of symptoms. Only transmural infarcts and LVFP > 12 mm Hg Durrer
Mortality cebo
(13 wk) 24%
for NP vs 13%
for pla-
Mortality cebo
(13 wk) 14% for NP vs 22%
for pla-
et alI9
328 patients, mean time of Rx 5 hours after onset of symptoms. All infarcts. (Patients with hypertension on admission were not excluded) CK-MB = MB isoenzyme
of creatine
Mortality (4 wk) 6% for NP vs 12% for placebo Peak CK-MB blood levels (anterior infarcts only)
kinase; LVFP = left ventricular
filling pressure;
NP = nitroprusside;
Hg was also required. Excluded were patients with normal left ventricular filling pressures, cardiogenie shock, hypertension requiring vasodilator therapy, severe bronchopulmonary disease or other serious illnesses, or a systolic blood pressure of < 100 mm Hg. Patients received either a 48-hour infusion of sodium nitroprusside or placebo. The mean time interval between the onset of symptoms and initiation of therapy was 17 hours. The 3-week mortality was 10.4% among the total population of placebotreated patients compared with 11.5% among all nitroprusside-treated patients (difference not significant). Likewise, after 13 weeks of follow-up, mortality rates were 19% in placebo-treated and 17% in nitroprusside-treated patients (difference not significant). No significant differences were noted in the peak CK-MB isoenzyme blood levels. However, with early treatment defined as therapy initiated
in Acute Myoqxudial John T. Flaherty,
In patll with acute myocardlal infarction, intravenous nitroglycerin lowers left ventricular filling pressure and systemic vascular resistance. At lower infuslon rates (46 pg/min) nitroglycerin is principally a venodilator, whereas at hiier infusion rates more balanced venous and arterial dilating effects are seen. Patients with left ventricular failure demonstrate increased or maintained stroke volumes, whereas patients without failure show a decrease in stroke volume. All hemodynamic subgroups show a decrease in left ventricular filling pressures and a reduction in electrocardiographic evidence of regional myocardiil ischemia. Longer-term infusions (20-48 hours) have been shown to result in myocardial preservation, as assessed by global and regional left ventricular functlln and laboratory indices of infarct s&e. Comparison of intravenous nitroglyercin and sodium nttroprusskle reveals increased intercoronary collateral flow with nitroglyercin, in contrast to a decrease with nitroprusskle, compatible with a “coronary steal.” Short-term administration of intravenous nttroglyercin with or without chronic administration of long-acting &rates have been found both to reduce short-term mortality and to have longterm beneficial effects on left ventricular remodeling in patients with anterior transmural infarctions. Current clinical practice woukl utilize intravenous nttroglyercin as inttiil therapy for patients receivhrg intravenous thrombolytic therapy and/or acute percutaneous transluminal coronary angioplasty within 4-6 hours of the onset of symptoms of acute myocardial infarction, in order to optimize intercoronary collateral flow until reperhrsion can be accompliihed. Patients reachhtg the hospital >6 hours but c 14 hours From the Division of Cardiology, Johns Hopkins of Medicine, Baltimore, Maryland. This article from Chapter 12, “The Role of Nitrates in Infarction and Post-Infarction,” by J.T. Flaherty, Pepine CJ, Thadani U, eds. Medical Therapy Disease: Nitrates, Beta Blockers, and Calcium ton: Little, Brown, 1992. Address for reprints: John T. Flaherty, Cardiology, Johns Hopkins University School more, Maryland 2120.5.
University School has been modified Acute Myocardial MD, in Abrams J, of Ischemic Heart Antagonists, BosMD, Division of of Medicine, Balti-
Infarction
MD
after symptom onset can still benefit from intravenous nttroglyercin for 2446 hours. Patients developi~ congestive heart failure and/or severe arterial hypertension as compliiins of ad acute i~rctlon as well as patients developing postinfarction unstable angina are also candidates for Intravenous nttroglyercin therapy. As wlth the a-ensin-converting enzyme inhibitors, r6 weeks of posthtfarction treatment with long-acting nltrates woukl also appear to reduce left ventricular remodeling in patients with anterior transmural infarctions. (Am J Cardiol166~70:726-6lB)
A
s recently as 1966 Friedberg’s Textbook of Cardiology warned against the adminiskation of nitroglycerin to patients with acute myocardial infarction. l Despite these warnings, cardiologists at the Johns Hopkins Hospital began studying the short-term hemodynamic and antiischemic effects of intravenous nitroglycerin in patients with acute myocardial infarction > 1.5 years ago. 2-5 More recently, we performed a randomized, placebo-controlled trial of intravenous nitroglycerin administered for 48 hours with thie goals of reducing infarct size and improving clinical outcomes.6 On October 1, 1981, intravenous nitroglycerin was released for general clinical use by the U.S. Food and Drug Administration. The clinical indications for which the intravenous preparation was approved include unstable angina and left ventricular failure complicating acute myocardial infarction. This article will review: (1) the acute hemodynamic responses to short-term infusion in patients with acute myocardial infarction; (2) the results of our placebo-controlled trial of long-term (48-hour) infusion of intravenous nitroglycerin in patients with acute myocardial infarction; (3) the results of other published clinical trials utilizing intraven0u.s nitroglycerin in an attempt to reduce infarct size; (4) a comparison of the hemodynamic and intercoronary collateral flow effects of nitroglycerin and nitroprusside; (5) the results of 2 placebo-con,A SYMPOSIUM:
NITROGLYCERIN
THERAPY
731B
trolled clinical trials of nitroprusside in acute myocardial infarction; and (6) the results of our study of the effects of acute administration of intravenous nitroglycerin in combination with intraaortic balloon counterpulsation therapy in patients with large transmural infarctions. Finally, I will present my own current recommendations for the use of intravenous nitroglycerin in patients with acute myocardial infarction. ACUTE HEMODYNAMIC EFFECTS OF SHORT-TERM INFUSION IN PAllENTS WITH ACUTE MYOCARDUU. INFARC’RON In 1972 cardiologists at the Johns Hopkins Hospital began administering intravenous mtroglycerin to patients with acute myocardial infarction with the goal of reducing regional myocardial ischemia and, when possible, improving left ventricular hemodynamics. In order to be sure that by lowering coronary perfusion pressure, regional myocardial ischemia was not being made worse, precordial ST-segment monitoring was employed. In the first 12 patients that we studied, only a small mean blood pressure decrease was induced (average -7 mm Hg), while at the same time a significant lowering of left ventricular filling pressure was documented.2 In all 12 cases a reduction in the sum of the precordial ST-segment voltages was observed.
Left Ventricular Filling (mm Hg)
Pressure
RGURE I Hemodynamk effects of intravenous nltroglycerln in the presence and absence of left ventdcular (LV) failure. Stroke volume Index (ml./mZ) Is plotted versus LV flllhlg pressure. llle responses obtained WlRl lntvavetlous Mroglycerln In bullvlduol hemodynamk subgroups are hdkated by arrows. A family of hypothetlcol Starllllg ventrkular timcUon cuwee Is IBed by doshed curves wlth dlmlnished contractllRy expressed by d ownward and rlghtward dlsplacemont. The vertical do+hed line ln&ates an upper llmlt of normal LV fllllng pressure. 746
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
70
Reassured by this initial experience, higher infusion rates of nitroglycerin were used in another clinical trial, which resulted in greater (15-30 mm Hg) reductions in mean arterial pressure.3 Beneficial anti-ischemic effects were observed in all 30 patients, irrespective of the presence or absence of left ventricular failure, as assessed by precordial ST-segment mapping. By stepwise titration of the intravenous infusion, nitroglycerin was shown to act principally as a venodilator at lower infusion rates, lowering left ventricular filling pressure by a mean of 10 mm Hg (45%), while causing minimal lowering of mean arterial pressure. At higher infusion rates, more balanced venous and arterial dilating effects were observed. In association with a similar lowering of left ventricular filling pressure (52%), a 20% lowering of mean arterial pressure was observed. Effects on left ventricular hemodynamics varied with the degree of underlying left ventricular failure (Figure 1). Patients without hemodynamic evidence of left ventricular failure demonstrated both a decrease in stroke volume and a decrease in the left ventricular filling pressure, evidencing the predominant preload lowering or “diuretic-like effect” induced by venodilation. In contrast, patients with mild left ventricular failure, as evidenced by an elevated left ventricular filling pressure but a normal stroke volume, demonstrated a lowering of left ventricular filling pressure but with a maintenance of stroke volume, suggesting that nitroglycerin was inducing a degree of afterload lowering as well. Patients with hemodynamic evidence of a more severe degree of left ventricular failure, as evidenced by both an elevated filling pressure and a reduced stroke volume, demonstrated the most beneficial hemodynamic effects. In this subgroup of patients, left ventricular filling pressure was again lowered, but in addition stroke volume was increased, compatible with a more balanced arterial and venodilating eflFect (Figure 1). Similar differential effects on hemodynamics have been reported for nitroprusside. This differential effect of nitroglycerin on stroke volume is most likely the result of differences in arterial impedance in patients with and without heart failure (Figure 2). Patients with moderate or severe left ventricular failure have been shown to have increased sympathetic tone, increased circulating catecholamines, and elevated systemic vascular resistance; the arterial-arteriolar actions of nitroglycerin are particularly useful in subjects with impaired left ventricular contractile function. SEPTEMBER
24, 1992
In summary, short-term infusion of intravenous nitroglycerin to patients with acute myocardial infarction resulted in improvement in electrocardiographic evidence of regional ischemia in all hemodynamic subgroups. Patients with the most severe degree of left ventricular failure demonstrated the most beneficial hemodynamic effects, obtaining both a reduction in pulmonary venous congestion and an improvement in forward cardiac output. Careful upward titration of the infusion rate allowed these beneficial anti-ischemic and hemodynamic effects to be obtained without excessive lowering of coronary pressure and without inducing a reflex tachycardia. In fact, the mean heart rate during intravenous nitroglycerin infusion was lower than control in all hemodynamic subgroups, most notably in the subgroups of patients with evidence of left ventricular failure. We also performed 2 additional acute hemodynamic studies in an effort to determine whether or not the anti-ischemic effects of intravenous nitroglycerin could be augmented by the addition of phenylephrine, which by maintaining coronary perfusion pressure might augment the anti-ischemic effects of nitroglycerin alone, or of propranolol, which by adding negative inotropic and negative chronotropic effects might augment the antiischemic effects of nitroglycerin by further reducing myocardial oxygen demand. The addition of the ol-adrenergic agonist and vasoconstrictor, phenylephrine, not only reversed the afterload-lowering effects, but also increased left ventricular filling pressure and reversed the beneficial anti-ischemic effects obtained with nitroglycerin alone.4 Addition of propranolol also resulted in a reversal of the preload-lowering effect obtained with nitroglycerin alone.5 Further, when the short-term nitroglycerin infusion was discontinued, left ventricular filling pressures rose above control levels, especially in patients with significant left ventricular dysfunction, reflecting the very short half-life of parenteral nitroglycerin (3 minutes) compared with the 6-hour elimination half-life of parenteral propranolol. RESULTS OF A RANdoMlZED PLACEBO-CONTROLLED CUNICAL TRW. OF LONG-TERM INFUSION OF NlTROGLYCERlN IN PATIENTS WITH ACm MYOCARDW INFARCTION
Based on the positive hemodynamic and antiischemic effects obtained on our initial short-term infusion studies,2,3 a randomized prospective trial of long-term infusion of nitroglycerin in patients with acute myocardial infarction was undertaken.6
Patients received a 48-hour infusion of eithe.r intravenous nitroglycerin or placebo, followed by the application of nitroglycerin or a placebo ointment for an additional 72 hours. Included in thi.s study were 104 patients with chest pain and electrocardiographic changes suggestive of acute myocardial infarction who presented < 12 hours after the onset of symptoms. Prior to initiation of the study drug infusion, we obtained a thallium-201 myocardial perfusion scan and a technetium-99m gated blood-pool scan. A computer-assisted scoring technique was used tie define a thallium defect score that provided a quantitative index of the extent of myocardium ait risk. It was recognized that the areas of decreaseld thallium uptake included both presently and previously infarcted myocardium as well as reversibly ischemic myocardium. The pretreatment gateld blood-pool scan (or, in a few patients, a 2-dimensional echocardiogram) defined baseline left ventricular function. The nitroglycerin infusion was begun at 5-10 pg/min, and the infusion rate increased stepwise until a 10% lowering of mean arterial pressure wa.s obtained. Blood pressure was monitored noninvalsively using an ultrasonic blood pressure cuff. Once this hemodynamic endpoint had been reached, the
I I
j
I
I
I I I I
I
, I I I
\ Severe
LV Dysfunction
INCREASING SYSTEMIC VASCULAR RESISTANCE
-
FIGURE 2. Dlfferentlal effects of afterload reduction on stroke volume according to hemcdynamlc subg~~p. Curves shown are for patients wRh normal left ventricular (LV) functkn as well as for patients wlth mlkl, moderate, and severe LV failure. Vertkal dashed llnes lndkzate pretmatmeht (right) and post-treatment (left) levels ofsystemk vascular resistance (SVR). Vertkal arrows lndkate the msignltude of the Increase In stroke volume lmlcrx that would be expected from this reduction In SVR for each hemodynamlc subgroup. A SYMPOSIUM:
NITROGLYCERIN
THERAPY
751B
final infusion rate was continued for 48 hours, unless side effects or hypotension required reduction or discontinuation of the infusion. The mean infusion rate required to reach the 10% mean blood pressure lowering endpoint was 90 pg/min. After 48 hours, the nitroglycerin infusion was discontinued, and 1 inch of nitroglycerin ointment was applied every 4 hours, with the dose increased stepwise, in an attempt to maintain the level of arterial pressure previously obtained with intravenous therapy. Control patients had 1 inch of placebo ointment applied every 4 hours. The thallium-201 gated blood-pool scans (or 2-dimensional echocardiogram) were obtained 7-14 days later. Of the 104 patients, 56 were randomized to intravenous nitroglycerin and 48 to placebo. There were no significant differences in admission clinical or scintigraphic parameters between the groups. Patients were divided into early- and late-treatment subgroups, according to the time interval between the onset of symptoms of acute infarction and time of the initiation of treatment. Patients treated < 10 hours from the onset of chest pain had “early” treatment, and those treated 210 hours after symptom onset had “late” treatment. The retrospective time cutoff of 10 hours divided the nitroglycerin-treated patients into 2 equal subgroups. Three-month mortality tended to be lower in the early nitroglycerin treatment subgroup (14%) compared with mortality in the other 3 treatment subgroups (21-28%). However, perhaps because of the small study population, this difference approached but did not reach statistical significance. When the unfavorable clinical outcomes of inhospital death, infarct extension, or development of new congestive heart failure were examined separately, the incidence of each tended to be lower with early nitroglycerin treatment, but again
PLACEBO 12 mm
10
47
48
LVFP
J CK infarct
SAP 10%
1 CK infarct size 37% but for inferior transmural infarcts only
SAP 20 mm Hg
J in-hospital mortality from 23% to 5% and precordial R waves better preserved in anterior infarcts
MAP 10%
Reduced infarct expansion, CK infarct size and & 3 month mortality
Inclusion Criteria
Patients (n) 104
60
Duration (hr)
85
Hg All infarcts
6
54
24
24
All infarcts
10
51
24-168
5
45
310
Only transmural
CK = creatine kinase; EF = ejection fraction; pressure; & = decrease.
IV = intravenous:
infarcts
LVFP = left ventricular
artery diastolic or pulmonary capillary wedge pressures of 2 12 mm Hg, transmural infarction, and an increase in total creatine kinase (CK) levels. Treatment was instituted a mean of 10 hours after the onset of symptoms, and nitroglycerin was titrated to lower the elevated pulmonary artery diastolic pressure; excessive lowering of arterial pressure was avoided. Total CK infarct size was found to be 23% lower for nitroglycerin-treated versus control patients (p 1.5 mm and a persistent R wave. A recent study by Jugdutt and WarnicalO reported beneficial effects of acute administration of intravenous nitroglycerin on early mortality as well as left ventricular ejection fraction, extent of left ventricular asynergy, and derived indices of infarct expansion and thinning. These data are discussed in Jugdutt’s review in this symposium. These investigators randomized 310 patients to intravenous nitroglycerin infusion or placebo for an average of 39 hours.ll Serial 2-dimensional echocardiography was employed to assess the incidence and severity of infarct expansion. In addition to demonstrating a reduction in expansion of the infarct segment when reassessed 10 days after admission, this large placebo-controlled trial was able to demonstrate significant reductions in CK infarct size and 3-month mortality, as well. The reduction in death was seen exclusively in the subjects with anterior wall infarction. Based on the premise that the majority of th.e published controlled clinical trials of intravenous A SYMPOSIUM:
NITROGLYCERIN
THERAPY
778
TABLE II Nitroprusside
Comparison in Acute
of Intravenous Myocardial
Nitroglycerin Infarction
with
Nitroglycerin Chairello
Nitroprusside
et alI3
Patients with anterior transmural myocardial infarction
j ST elevations
Open chest anesthetized dog model with acute coronary ligation
& ST elevations f MBF
Mann et alI4 Patients with angiographically visible intercoronary channels MBF = myocardial
t ST elevations
and
t MBF
t ST elevations 1 MBF
and
1 MBF
blood flow; 1 = decrease;
t = increase.
nitroglycerin in acute myocardial infarction failed to detect an unequivocal reduction in mortality due to their relatively small sample sizes, Yusuf and Collinsl* pooled mortality data from 7 randomized placebo-controlled studies. Although only 2 of the 7 trials9J1 were able to demonstrate a statistically significant reduction in mortality, pooling the mortality data from all 7 studies, which included > 1,000 patients, revealed a highly significant 30% reduction in mortality in the intravenous nitroglycerin-treated patients versus control patients. COMPARISON OF THE EFFECTS OF NITROGLYCERIN AND NlTROPRUSSlDE ON REGIONAL ISCHEMlA AND INTERCORONARY COLLATERAL BLOOD FLOW
Several experimental animal and clinical studies have compared the hemodynamic and anti-ischemit effects of nitroglycerin and nitroprusside (Table II). These suggest that nitroprusside may induce a coronary steal phenomenon with subsequently differing effects on the myocardium, compared with beneficial anti-ischemic actions of nitroglycerin.13-l6 In order to define the mechanisms responsible for the apparent opposite effects of nitroglycerin and nitroprusside on regional ischemia, Chiarello et all3 carried out a study in an open chest canine model. Following ligation of the left anterior descending coronary artery, nitroprusside or nitroglycerin were titrated to lower mean arterial pressure by 200 mm Hg. Utilizing the radioactive microsphere technique, myocardial blood flow distal to the ligated left anterior descending coronary artery was found to decrease and epicardial ST-segment elevations to increase during nitroprusside infusion. In contrast, during nitroglycerin infusion, ischemic zone myocardial blood flow increased and 78B
THE
AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
70
epicardial ST segment elevations decreased. Since the coronary artery remained ligated, the improvement in regional ischemia observed with nitroglycerin must be taken as evidence of an increase in intercoronary collateral blood flow. In contrast, during nitroprusside infusion, collateral flow to the ischemic zone actually decreased, and flow to the nonischemic zone increased, providing evidence of a coronary steal. In the cardiac catheterization laboratory, Mann et all4 used the xenon-133 washout technique to study the effects of nitroglycerin and nitroprusside on regional myocardial blood flow. In a subgroup of patients with fixed coronary artery disease and angiographically visible intercoronary collaterals, myocardial blood flow distal to a severe coronary artery stenosis was found to increase with nitroglycerin and to decrease with nitroprusside. Cappuro et a1,15using an open chest anesthetized dog model with well-developed collateral vessels induced by previous implantation of an amaroid constrictor, demonstrated greater sensitivity of intercoronary collateral channels to the vasodilating effects of nitroglycerin than of nitroprusside. Finally, Marcho and Vatner16 demonstrated in a conscious, instrumented dog model that, compared with nitroprusside, nitroglycerin induces relatively less dilation of the small resistance vessels and greater dilation of the large (conductance) coronary vessels, which include intercoronary collateral channels. CUNICAL TRlALS OF NCTROPRUSSIDE ACUTE MYOCARDW INFARCTlON
IN
In 2 published prospective randomized, placebocontrolled trials, nitroprusside or a placebo was administered to patients with acute myocardial infarction in an attempt to reduce infarct size and/or improve short-term mortality (Table III). In a large multicenter Veterans Administration Cooperative Trial, nitroprusside was titrated to reach 1 of 4 endpoints: (1) a reduction of left ventricular filling pressure to < 60% of control; (2) a lowering of systolic arterial pressure to 20% of the control pressure plus 76 mm Hg; (3) the development of significant side effects; or (4) a maximum infusion rate of 200 pg/min.18 Included were 812 patients with the onset of chest pain < 24 hours prior to admission, and transmural infarction, as evidenced by ST-segment elevations and subsequent development of pathologic Q waves or a new conduction defect (Table III). Approximately 67% of patients had anterior wall infarctions. A left ventricular filling pressure of r 12 mm SEPTEMBER
24,
1992
TABLE
III Clinical
Trials of Intravenous
Nitroprusside
in Acute Myocardial
Clinical Trial
Infarction Late Treatment Group
Early Treatment Group
Cohn et alI8 812 patients, mean time of Rx 17 hours after onset of symptoms. Only transmural infarcts and LVFP > 12 mm Hg Durrer
Mortality cebo
(13 wk) 24%
for NP vs 13%
for pla-
Mortality cebo
(13 wk) 14% for NP vs 22%
for pla-
et alI9
328 patients, mean time of Rx 5 hours after onset of symptoms. All infarcts. (Patients with hypertension on admission were not excluded) CK-MB = MB isoenzyme
of creatine
Mortality (4 wk) 6% for NP vs 12% for placebo Peak CK-MB blood levels (anterior infarcts only)
kinase; LVFP = left ventricular
filling pressure;
NP = nitroprusside;
Hg was also required. Excluded were patients with normal left ventricular filling pressures, cardiogenie shock, hypertension requiring vasodilator therapy, severe bronchopulmonary disease or other serious illnesses, or a systolic blood pressure of < 100 mm Hg. Patients received either a 48-hour infusion of sodium nitroprusside or placebo. The mean time interval between the onset of symptoms and initiation of therapy was 17 hours. The 3-week mortality was 10.4% among the total population of placebotreated patients compared with 11.5% among all nitroprusside-treated patients (difference not significant). Likewise, after 13 weeks of follow-up, mortality rates were 19% in placebo-treated and 17% in nitroprusside-treated patients (difference not significant). No significant differences were noted in the peak CK-MB isoenzyme blood levels. However, with early treatment defined as therapy initiated
