Time course of direct cardiotoxic effects of high cocaine concentration in isolated rabbit heart.

Studies of acute cocaine cardiotoxicity are often confounded by the effects of cocaine on peripheral hemodynamics and an intact sympathetic nervous system. To study the direct effects of an acute dose of cocaine on heart, we used Langendorff-perfused isolated rabbit hearts. The hearts were attached to a Langendorff perfusion system through the aorta and were perfused with Krebs-Henseleit buffer. In nonpaced hearts, cocaine in concentrations of 10, 25, and 50 microM (reported to be in the range of drug concentrations lethal for humans) caused dose-dependent deterioration in heart contractility (decrease in +dP/dt, -dP/dt, and developed pressure) and coronary flow, when measured at 10, 20, and 30 min of the protocol. Cocaine 25 and 50 microM also caused a decrease in heart rate (HR). In paced hearts treated with cocaine 50 microM, early events included a progressive decrease in +dP/dt (by 15% vs. baseline at 20 s of perfusion and by 26% vs. baseline and 19% vs. control group at 30 s of perfusion; p < 0.05) and in -dP/dt (by 20% vs. baseline at 30 s of perfusion and by 25% vs. baseline and 15% vs. control group at 40 s of perfusion, p < 0.05). Decrease in HR (failure to pace) was observed only at 60 s of perfusion and averaged 20% versus both baseline and control group (p < 0.05). No changes in coronary flow were observed during the first 60 s after onset of cocaine administration.(ABSTRACT TRUNCATED AT 250 WORDS)

Absence of hemodynamic deterioration in the presence of amlodipine following experimental myocardial infarction.

In general, calcium channel blockers have not been used in patients during acute myocardial infarction as they may exacerbate heart failure, possibly by neuro-humoral stimulation. Amlodipine, a new dihydropyridine calcium channel blocker without neurohumoral stimulation, was tested in an experimental model of acute myocardial infarction with assessment of hemodynamics, left ventricular (LV) function, and infarct size. Anesthetized dogs were subjected to 3 h of coronary artery occlusion followed by 3 h of reperfusion. At 2 h of occlusion, they were randomized to receive either amlodipine (250 micrograms/kg, n = 11) or saline (n = 11). Before treatment, all variables were similar in both groups. The diastolic pressure was unchanged following saline, but was reduced following amlodipine by 1 h after therapy (from 94 +/- 5 to 71 +/- 3 mm Hg, p < 0.0001 vs. saline) and for the duration of the protocol. Indices of left ventricular (LV) function did not deteriorate with amlodipine treatment compared with saline. After 3 h of reperfusion, the LV dP/dt was 1,720 +/- 114 mm Hg/s in the saline group and 1,958 +/- 167 mm Hg/s with amlodipine (p = ns). The area ejection fraction, assessed by echocardiography, was similar in both groups (43 +/- 5%, saline; 45 +/- 3%, amlodipine; p = ns), as was the LV end-diastolic pressure (8 +/- 1 mm Hg, saline; 7 +/- 1 mm Hg, amlodipine; p = ns). Subendocardial regional myocardial blood flow, measured by radioactive microspheres, was 0.75 +/- 0.08 ml/min/g with saline and 1.34 +/- 0.33 ml/min/g with amlodipine in the previously ischemic reperfused subendocardium (p = 0.1).(ABSTRACT TRUNCATED AT 250 WORDS)

Influx of neutrophils into the walls of large epicardial coronary arteries in response to ischemia/reperfusion.

BACKGROUND: There are several clinical situations in which large epicardial coronary arteries are deprived of blood flow, such as occurs when an obstructing thrombus or embolus lodges within a vessel or during coronary dissection. There is little information concerning the effect of flow deprivation on large epicardial coronary arteries. METHODS AND RESULTS: We studied a model in which a segment of a large epicardial coronary artery was deprived of blood flow using both proximal and distal clamps for 3 hours followed by reperfusion. On examination by light microscopy of cross sections of the arteries, 19 +/- 6 neutrophils were present in the intima of ischemic/reperfused vessels, whereas only a mean of 4 +/- 3 (SEM) were present in the intima of nonischemic vessels (p less than 0.02). On average, there were 17 +/- 9 neutrophils just under the elastic lamina in ischemic/reperfused vessels versus none in the nonischemic vessels (p less than 0.05); there were 16 +/- 10 neutrophils present within the media of ischemic/reperfused vessels, and none (p less than 0.05) in the nonischemic vessels. Electron microscopic analysis revealed that neutrophils in the ischemic/reperfused vessels were often "sandwiched" between the endothelial cells and the elastic lamina. Ultrastructural abnormalities within the myocardium also revealed damage to the microvasculature, including the presence of neutrophils within the vessels and erythrocyte stasis. To rule out the possibility that findings in the large epicardial arteries were due to toxic substances from static blood within the isolated arterial segment, a protocol was performed in which blood was removed from the isolated segment. Again, neutrophil infiltration into the vessel was observed. Resting mean epicardial coronary artery blood flow before coronary occlusion was 19 +/- 3 ml/min; mean coronary blood flow 2.5 hours after reperfusion was identical at 19 +/- 3 ml/min. Response to both endothelial-dependent vasodilation (acetylcholine) and endothelial-independent vasodilation (nitroglycerin) challenges was normal early after reperfusion but was depressed late after reperfusion, suggesting progressive vascular dysfunction and hence a form of vascular reperfusion injury in this model. CONCLUSIONS: When large epicardial coronary arteries are deprived of blood flow, followed by reperfusion in this model, neutrophils migrate into the vessel wall as well as into the microvasculature. These abnormalities are associated with reduced endothelial-dependent and endothelial-independent coronary vasodilator reserve.

Nifedipine protects the heart from the acute deleterious effects of cocaine if administered before but not after cocaine.

BACKGROUND: We tested the hypothesis that nifedipine, a calcium channel blocker, could ameliorate the toxic effects of cocaine on the myocardium. METHODS AND RESULTS: In an initial protocol, anesthetized dogs were pretreated with nifedipine or saline and then administered cocaine (10 mg/kg, i.v. bolus). Coronary blood flow, heart rate, mean arterial pressure, and the first derivation of left ventricular pressure (dP/dt) were measured at baseline, 2 minutes, and 15 minutes after cocaine administration. Nifedipine pretreatment prevented the early cocaine-induced decrease in coronary blood flow and improved left ventricular dP/dt compared with untreated control animals. After cocaine, ejection fraction fell in the saline group to 37 +/- 3% but increased in the nifedipine group to 59 +/- 4% (p less than 0.05). In a second protocol, vehicle or intravenous nifedipine was administered after an infusion of cocaine (10 mg/kg). In contrast to pretreatment, there was no significant improvement in left ventricular function or coronary blood flow in nifedipine-treated versus control animals. Data from the study also suggested that cocaine acts directly on the myocardium. Within seconds of cocaine bolus administration, coronary blood flow in control animals increased to a peak level 59 +/- 14% higher than before cocaine and left ventricular dP/dt decreased by 23 +/- 5%, providing evidence that cocaine causes direct depression of myocardial function independent of a decrease in myocardial blood flow. CONCLUSIONS: We conclude that nifedipine administered as a pretreatment protects against the depression of myocardial function and decrease in coronary blood flow caused by acute cocaine administration. However, when nifedipine is given after cocaine, no improvement is seen. Cocaine has a direct negative inotropic effect on the heart that is independent of a decrease in coronary blood flow.

Adverse effects of cocaine on cardiovascular dynamics, myocardial blood flow, and coronary artery diameter in an experimental model.

Smoking "crack" is currently popular; this results in rapidly peaking, high blood levels of cocaine. Our purpose was to investigate the effects of rapid cocaine administration. Hemodynamics, myocardial blood flow, and left ventricular cavity end-systolic and end-diastolic areas were measured in pentobarbital-anesthetized dogs before and 15 minutes after (1) a bolus intravenous injection of cocaine (n = 6), or (2) a 10-minute infusion of cocaine (n = 6), or (3) saline (n = 6) administration. In both treated groups cocaine caused a significant reduction in heart rate and left ventricular dP/dt compared with baseline measurements, and an increase in left ventricular end-diastolic pressure. Cocaine also caused dilation of the left ventricle and a fall in regional myocardial blood flow. In nine other dogs, proximal circumflex artery diameter, assessed by angiography, and regional myocardial blood flow were measured before and 3 to 5 minutes after cocaine (n = 7) or saline (n = 2). In treated animals, circumflex artery diameter was reduced 15 +/- 4% (range 2% to 29%, p less than 0.01 versus baseline) after cocaine. We conclude that in this model, rapid administration of cocaine depresses left ventricular function, causes left ventricular dilation, and is associated with coronary artery vasoconstriction and reduced myocardial blood flow.

Evaluation of nonradioactive, colored microspheres for measurement of regional myocardial blood flow in dogs.

Measurement of regional myocardial blood flow (RMBF) is crucial in experimental studies of myocardial ischemia and reperfusion in dogs. The standard measurement technique uses radioactive microspheres; however, not all institutions are able to dispose of radioactive waste and therefore cannot make use of this method. We tested a new, nonradioactive microsphere, labeled with colors instead of nuclides. Simultaneous blood flow measurements with two nuclide-labeled and two colored microspheres were performed after coronary occlusion in dogs. Both techniques show a within-method correlation of r greater than 0.98. Duplicate variability for paired RMBF values in 80 samples was 8.7 +/- 0.1% when computed with radioactive microspheres and 13.2 +/- 1.8% when computed with colored microspheres. There was a good correlation in the measurement of RMBF between the radioactive- and colored-microsphere methods (r = 0.98). The best-fitting linear regression line was expressed by the formula: Colored-microsphere RMBF = 1.11 (radioactive-microsphere RMBF)-0.02. When measured by colored microspheres, RMBF was approximately 8% higher than when computed with radioactive microspheres for blood flow values of 0-2 ml/min/g. When blood flow was increased pharmacologically to levels of 2-7.5 ml/min/g, colored microspheres yielded blood flow values 39% higher than the values computed by radioactive microspheres. We conclude that the nonradioactive, colored-microsphere method correlates with the radioactive technique, but at high flows, it yields values greater than those obtained with radioactive microspheres.

Functional infarct expansion, left ventricular dilation and isovolumic relaxation time after coronary occlusion: a two-dimensional echocardiographic study.

Left ventricular dilation and infarct expansion after acute myocardial infarction are associated with an increased morbidity and mortality. The purpose of this study was to determine whether angiotensin-converting enzyme inhibition could reverse left ventricular dilation and improve the diastolic properties of the left ventricle very early after coronary occlusion. The acute time course of left ventricular dilation and infarct expansion (as determined by two-dimensional echocardiography) and early diastolic isovolumic relaxation time were studied in 20 dogs subjected to 3 h of coronary occlusion. End-diastolic area before occlusion was 8.4 +/- 0.5 and 8.9 +/- 0.7 cm2 (p = NS) in the captopril- and the saline-treated group, respectively. At 30 min after occlusion (pretreatment), end-diastolic area increased to 12.6 +/- 0.8 cm2 in the captopril-treated group (p less than 0.01) and 11.3 +/- 0.9 cm2 (p less than 0.05) in the saline-treated group. Three hours after occlusion and after captopril treatment, end-diastolic area decreased to 9.4 +/- 0.6 cm2 (p less than 0.05 versus 30 min after occlusion), whereas it was unchanged in the saline-treated group. Functional infarct expansion (as assessed by end-systolic anterior to posterior endocardial segment length ratio) occurred early after occlusion, and captopril reduced this expansion. Pretreatment values for early diastolic isovolumic relaxation time increased from 29.1 +/- 2.4 to 50.5 +/- 2.9 ms in captopril-treated dogs (p less than 0.01) and from 34.3 +/- 3.4 to 46.9 +/- 2.7 ms in saline-treated dogs (p less than 0.01) after coronary occlusion, implying a worsening of diastolic function.(ABSTRACT TRUNCATED AT 250 WORDS)

The effect of digitalis on experimental myocardial infarct size and hemodynamics.

Early studies suggested that digitalis exacerbated ischemia (ST segment data); however, there are no studies assessing the effect of this agent on anatomic infarct size with the use of a risk zone technique. Therefore, the aim of this study was to assess quantitatively whether digitalis extends necrosis in a model of coronary artery occlusion. Anesthetized dogs were subjected to 6-hour occlusion and, 30 minutes after occlusion, were randomized to digoxin (250 micrograms bolus/5 min intravenously, n = 9) or saline (n = 9) groups. At 6 hours, in vivo area at risk was determined by monastral blue dye injection and area of necrosis was assessed by tetrazolium staining. Heart rate and blood pressure were not different between groups before treatment or at 6 hours after occlusion. Left ventricular dP/dt was similar in both groups before occlusion (2350 +/- 293 mm Hg/sec digoxin vs 1839 +/- 122 mm Hg/sec saline, p = NS), but after 6 hours of coronary occlusion, had increased in the digoxin group to 2583 +/- 340 mm Hg/sec while it decreased in the saline group to 1517 +/- 128 mm Hg/sec (p less than 0.05 between groups at 6 hours), suggesting that digoxin increased contractility. Area at risk was 17.7 +/- 1.3% of the left ventricle in the digoxin group and 20.9 +/- 2.0% of the left ventricle in the saline group (p = NS). Area of necrosis, expressed as a percentage of area at risk, was 90.0 +/- 3.5% in the digoxin group vs 88.6 +/- 2.1% in the saline group (p = NS). Therefore, during acute myocardial infarction, digitalis confers a moderate increase in contractility without extending necrotic damage.

Effect of verapamil on infarct size in dogs subjected to coronary artery occlusion with transient reperfusion.

Reocclusion after successful coronary reperfusion occurs in 15 to 35% of patients receiving thrombolytic therapy for acute myocardial infarction. The present study was designed to simulate the clinical situation of reocclusion and determine whether verapamil might be effective in reducing myocardial necrosis and preserving high energy phosphates in this setting. Pentobarbital-anesthetized, open chest dogs underwent occlusion of the left anterior descending coronary artery for 2 hours followed by 1 hour of reperfusion and a further 4 hours of coronary artery occlusion. Treatment with verapamil (intravenous bolus dose of 0.2 mg/kg body weight followed by infusion of 0.56 +/- 0.14 mg/kg per h) was begun 1 hour after occlusion and infusion was continued for the remainder of the experiment. The dose of verapamil was adjusted to lower mean arterial pressure to approximately 90 mm Hg. The area at risk was determined by intraatrial injection of monastral blue dye and the area of necrosis was assessed by triphenyltetrazolium chloride staining. In vivo myocardial needle biopsy for determination of adenosine triphosphate and creatine phosphate was performed at the end of the experiment. The area of the left ventricle at risk was similar in both groups (control [n = 8], 20.2 +/- 1.6% versus verapamil-treated [n = 9], 23.1 +/- 2.9%; p = NS). The area of necrosis expressed as a percent of the area at risk was reduced in the verapamil-treated group compared with the control group (43.3 +/- 5.0% versus 63.1 +/- 6.8%, p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Alterations in the distribution of high-energy phosphates during ischemia in a canine model of reperfusion-induced ventricular fibrillation.

High-energy phosphate metabolites of the canine heart were analyzed before coronary artery occlusion and after 15 minutes of ischemia, and the results were then correlated with the occurrence of ventricular fibrillation upon reperfusion (RVF). Animals which developed VF upon reperfusion after 15 minutes of ischemia had lower levels of creatine phosphate and endocardial adenosine triphosphate (ATP), and increased accumulation of the catabolites of ATP metabolism, inosine and hypoxanthine. Animals which developed RVF also had lower levels of regional myocardial blood flow in the center of the ischemic zone during the period of coronary occlusion. Occluded bed size was the same in dogs which did and did not develop RVF. These data suggest that VF upon reflow is associated with more severe ischemia and an increase in high-energy phosphate catabolism during the period of ischemia.

Effects of transient increased afterload during experimentally induced acute myocardial infarction in dogs.

Alterations in afterload may occur during acute myocardial infarction (AMI), but it is unknown whether such alterations cause long-term changes in the left ventricular topography or alter healing of the AMI. AMI was produced by ligation of the left anterior descending coronary artery in open-chest dogs. Eight dogs were randomized to a methoxamine group with an infusion dose of 30 micrograms/kg/min starting 1 hour after ligation for 4 hours to increase systemic systolic pressure by 40 to 50 mm Hg, and 8 were randomized to a saline control group (n = 8). Seven days later the dogs were killed and the hearts examined. The ratio of infarct wall thickness to noninfarct wall thickness was 1.13 +/- 0.03 (mean +/- standard error of the mean) in control dogs and was 0.98 +/- 0.03 in the dogs treated with methoxamine (p less than 0.005). An expansion index was determined as previously reported and expansion was considered to have occurred if this index exceeded 1.09. The expansion index was 0.98 +/- 0.06 in the control group and 1.18 +/- 0.07 in the methoxamine group (p less than 0.05). Histologic analysis suggested a lag in the healing rate in the methoxamine-treated dogs. Thus, early, brief increases in afterload cause infarct expansion and thinning and appears to slow the early healing phase of AMI in dogs.

Preservation of high-energy phosphates by verapamil in reperfused myocardium.

To determine whether verapamil prevents depletion of adenine nucleotides during and after severe myocardial ischemia, dogs were subjected to 15 min occlusions of the left anterior descending coronary artery followed by 240 min of reperfusion. One hour before occlusion, dogs were randomly assigned to a treatment group (n = 10) to which an infusion of intravenous verapamil was given until the onset of reperfusion or to an untreated saline group (n = 9). Verapamil reduced mean aortic pressure and heart rate. After 15 min of ischemia, endocardial adenosine triphosphate (ATP) level, determined by needle biopsy, decreased in the untreated group from 34.7 +/- 2.0 to 24.4 +/- 2.7 nmol X mg protein-1 (p less than .005 vs preocclusion) and in the verapamil group from 32.8 +/- 1.5 to 30.3 +/- 1.5 nmol X mg protein-1 (NS vs preocclusion). Dogs receiving verapamil had significantly higher ATP levels than untreated animals after 90 and 240 min of reperfusion. In untreated animals the sum of inosine and hypoxanthine levels increased during occlusion from very low levels to 4.6 +/- 1.1 nmol X mg protein-1 in the epicardium and to 6.8 +/- 1.5 nmol X mg protein-1 in the endocardium (p less than .05 compared with preocclusion values). In verapamil-treated dogs inosine and hypoxanthine levels increased to only 1.2 +/- 0.3 (epicardium) and 1.9 +/- 0.6 nmol X mg protein-1 (endocardium) (both NS compared with preocclusion values). After 90 min of reperfusion the sum of ATP, adenosine diphosphate, adenosine monophosphate, inosine, and hypoxanthine levels was decreased in the endocardium by 10.2 nmol X mg protein-1 in the untreated group, but no change was observed in verapamil-treated animals. We conclude that breakdown of ATP to inosine and hypoxanthine during severe ischemia is reduced by verapamil, resulting in higher ATP concentrations during occlusion and reperfusion and decreased washout of the diffusible purines inosine and hypoxanthine during reperfusion.

The effect of streptokinase on intramyocardial hemorrhage, infarct size, and the no-reflow phenomenon during coronary reperfusion.

The purpose of this study was to determine whether streptokinase exacerbates intramyocardial hemorrhage during coronary reperfusion, has any intrinsic effect on myocardial infarct size other than its ability to lyse proximal thrombi in coronary arteries, and can abolish the no-reflow phenomenon. Anesthetized open-chest dogs underwent coronary occlusion for 3 hr followed by 3 hr of reperfusion. Area of infarct was assessed by tetrazolium staining, anatomic zone of no-reflow by injection of the fluorescent dye thioflavin S at the end of the reperfusion period, regional blood flow during occlusion and reperfusion by the radioactive microsphere technique, and extent of gross hemorrhage by assessment of photographic enlargements of the heart slices. Area of infarction of the left ventricle was similar in control (13.4 +/- 3.6%) and streptokinase-treated dogs (13.0 +/- 2.9%; p = NS). Seven of eight dogs in the untreated group had anatomic perfusion defects as assessed by thioflavin S at the end of the reperfusion phase; seven of eight dogs in the streptokinase group had anatomic perfusion defects. There was no difference in the extent of gross hemorrhage between the two groups (6.5 +/- 2.1% of left ventricle in controls and 5.7 +/- 2.3% in streptokinase-treated dogs). Severe depression of regional blood flow during reperfusion was present within the infarcted tissue and was associated with an anatomic perfusion defect as defined by thioflavin S; there was moderate depression of flow within the noninfarcted, salvaged subepicardium. In a separate series of experiments, infarcts were assessed for hemoglobin content. Intramyocardial hemoglobin levels were not higher after fibrinolytic therapy plus reperfusion compared with reperfusion alone.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of recurrent ischemia on myocardial high energy phosphate content in canine hearts.

Successive ischemic episodes are commonly seen in clinical and experimental cardiology. Although prolonged abnormalities of the canine myocardium have been described following a single ischemic episode, it is not known whether myocardial injury is cumulative following multiple ischemic episodes. Dogs were subjected to three 15-min left anterior descending coronary artery occlusions, each followed by 30 min of reperfusion. In vivo biopsies were obtained for biochemical analysis before and during the first occlusion and 30 min into each reperfusion period. ATP and creatine phosphate (CP) fell in the endocardium during occlusion by 24% and by 69% respectively (both p less than .0001). While CP recovered during each reperfusion period, ATP remained significantly depressed. Loss of membrane-permeable purine nucleotide synthesis precursors occurred with the first reperfusion period but not with the second and third reperfusion periods. Therefore, reperfusion following one 15-min coronary occlusion is associated with severe depression of myocardial high energy phosphates; however, two additional occlusions do not cause a further decrease of these substances when intermittent reperfusion is allowed for 30 min.

Correlates of reperfusion ventricular fibrillation in dogs.

To elucidate determinants of reperfusion ventricular fibrillation (VF), regional myocardial blood flow, ATP, creatine phosphate (CP), heart rate and blood pressure were compared in 2 groups of anesthetized dogs: those that fibrillated spontaneously upon release of a 15-minute coronary artery occlusion (VF group, n = 8) and those that did not fibrillate when reperfused (No VF group, n = 27). Arterial pressure and heart rate before and during coronary artery occlusion were similar in both groups. Ischemic endo- and epicardial ATP values, measured at the end of the occlusion period, were reduced approximately 20% of nonischemic values in both groups. In contrast, CP (nmol . mg protein-1) within the ischemic zone was significantly lower in the VF group in both the epicardium (14.3 +/- 1.6 in the VF group vs 22.8 +/- 2.5 in the No VF group, p less than 0.01) and the endocardium (9.0 +/- 2.0 in the VF group vs 18.7 +/- 1.8 in the No VF group, p less than 0.01). Furthermore, epi- and endocardial regional myocardial blood flow in the center of the ischemic zone during occlusion was significantly lower in VF dogs than in No VF dogs. Epicardial flow was 0.06 +/- 0.03 ml X min-1 X g-1 in VF dogs vs 0.44 +/- 0.06 in No VF dogs (p less than 0.001) and endocardial flow was 0.03 +/- 0.02 ml X min-1 X g-1 in VF dogs vs 0.23 +/- 0.04 ml X min-1 X g-1 in No VF dogs (p less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)

Morphologic and functional effects of piroxicam on myocardial scar formation after coronary occlusion in dogs.

To determine whether piroxicam, a widely used, long-acting anti-inflammatory agent, causes scar thinning after acute myocardial infarction (MI), MI was produced in 16 anesthetized, open-chest dogs by ligation of the proximal left anterior descending coronary artery. The dogs were randomized into 2 groups and treated in a blinded fashion, 8 with piroxicam, 1 mg/kg i.v. at 15 minutes and at 3 hours after ligation (Group 1) and 8 with saline solution (Group 2). Two-dimensional echocardiograms were performed 7 days and 6 weeks after ligation. At 6 weeks, the dogs were killed and the hearts examined. Scar thickness was 7.1 +/- 0.3 mm in control dogs and 5.2 +/- 0.4 mm in piroxicam-treated dogs (p less than 0.01). The ratio of scar thickness to noninfarcted wall thickness was 0.87 +/- 0.03 (mean +/- standard error of the mean) in the control group, and was significantly lower (0.62 +/- 0.04) in the piroxicam-treated group (p less than 0.001). Regional function, expressed as the percent change in the area of the left ventricular cavity (% delta A) from short-axis 2-dimensional echocardiograms, was 42 +/- 3% 7 days after occlusion in the control group and was not significantly different in the treated group (34 +/- 5%). At the end of 6 weeks % delta A had improved in the piroxicam-treated group to 44 +/- 3% (p less than 0.05 compared with the value after 7 days), and was similar to % delta A of the control group at 6 weeks (43 +/- 3%). Thus, clinical doses of piroxicam administered early after MI caused moderate scar thinning, which was not associated with impairment of regional left ventricular function 6 weeks later.