The occurrence, outcome, and predictors of myocardial infarct extension were determined in 848 patients with acute myocardial infarction. An increase in the level of plasma MB creatine kinase activity was used to detect extension, which occurred in 71 of 848 patients (8.4%). For these patients, hospital mortality was more than four times higher than for those without extension (30% versus 7%, P less than 0.01). However, for patients surviving the initial hospitalization, there was no significant difference in mortality during the following year (12% compared with 9%). Multivariable analyses indicated that extension was more likely to occur in patients with recurrent ischemic pain during the second hospital day, a history of previous myocardial infarction, and ST segment depression on the admission electrocardiogram. The occurrence of extension in patients with two of these risk factors was more than twice that of patients without any of the risk factors (15.1% compared with 5.8%). Patients with these risk factors should be considered for early coronary angiography and possible intervention to prevent infarct extension and its sequellae.

A multicentred, randomised, blind study was started in 1978 to compare propranolol or hyaluronidase with placebo in patients with acute myocardial infarction admitted within 18 hours of onset of symptoms. Patients were randomised to group A and received hyaluronidase, propranolol, or placebo, or, if propranolol was contraindicated, to group B and received hyaluronidase or placebo. Hyaluronidase (500 U/kg given every six hours for 48 hours) had no effect on mortality or infarct size in the overall population. Because spontaneous reperfusion was more common in patients with early peaking of plasma creatine kinase MB or non-transmural electrocardiographic changes or both, the results were reanalysed for two subgroups: those in whom plasma creatine kinase peaked less than 15 hours after the onset of symptoms (early peak, n = 184) and those with a peak greater than 15 h after the onset of symptoms (late peak, n = 546). The distribution of time to peak activity of creatine kinase MB was similar in the hyaluronidase and placebo groups. In the early peak patients who were given hyaluronidase (groups A and B) total mortality and cardiac-specific four year mortality were significantly lower. This was most pronounced in group B in which the total mortality was 45% and cardiovascular mortality was 47% less than in the placebo group. Similarly, mortality from cardiovascular disease in patients (groups A and B) with nontransmural ischaemia (ST-T changes) given hyaluronidase was significantly lower, with group B showing a 50% reduction. In the subsets of patients with late peaking of creatine kinase MB or those presenting with transmural electrocardiographic changes there was no difference in total mortality or deaths from cardiac disease between those given hyaluronidase and those given placebo. Hyaluronidase was associated with improved survival in patients with early peaking of plasma creatine kinase MB, suggesting the possibility of salvage of myocardium in patients who have early spontaneous reperfusion and possibly after therapeutic reperfusion.

Data were analyzed from 698 patients with proved acute myocardial infarction (AMI) to develop a method to predict the occurrence of complete heart block (CHB). The presence of electrocardiographic abnormalities of atrioventricular or intraventricular conduction during hospitalization was determined for each patient. The electrocardiographic risk factors considered were: first-degree atrioventricular block, Mobitz type I atrioventricular block, Mobitz type II atrioventricular block, left anterior hemiblock, left posterior hemiblock, right bundle branch block and left bundle branch block. A CHB risk score was developed that consisted of the sum of each patient's individual risk factors. CHB risk scores of 0, 1, 2 or 3 or more were associated with incidences of CHB of 1.2, 7.8, 25.0 and 36.4%, respectively. When applied to an independent AMI data base, as well as to the summed results of 6 previously reported series that identified predictors of CHB during AMI, a similar incremental risk of CHB as predicted by the risk score method was demonstrated.

Recent studies have led to controversy about whether long-term digoxin therapy after confirmed or suspected myocardial infarction increases mortality. We analyzed the mortality experience in 903 patients enrolled in the Multicenter Investigation of Limitation of Infarct Size (MILIS). As in previous studies, the decision to treat or not to treat with digoxin was made by the patient's personal physician on the basis of the usual clinical indications. Cumulative mortality was 28 percent for the 281 digoxin-treated patients as compared with 11 percent for the 622 patients who did not receive digoxin (P less than 0.001; follow-up interval, six days to 36 months; mean, 25.1 months). However, patients treated with digoxin had more base-line characteristics predictive of mortality than did their counterparts. Adjustment for these differences with two separate applications of the Cox method yielded P values of 0.14 and 0.34 for tests of difference in mortality, providing no evidence for a significant excess mortality associated with digoxin. Thus, the findings in the MILIS population do not support the assertion that digoxin therapy is excessively hazardous after infarction, but the existence of an undetected harmful effect can only be excluded with a randomized study. Until the results of such a study are available, we recommend careful consideration of whether any treatment of ventricular dysfunction is actually needed, consideration of alternatives to digoxin therapy, and restriction of digoxin use to the subgroup of patients (with severe chronic congestive failure and a dilated left ventricle) previously shown to have a beneficial clinical response.

The effects of abrupt withdrawal or continuation of beta-blockade therapy during acute myocardial infarction were evaluated in 326 patients participating in the Multicenter Investigation of the Limitation of Infarct Size (MILIS). Thirty-nine patients previously receiving a beta-blocker and randomly selected for withdrawal of beta-blockers and placebo treatment during infarction (group 1) were compared with 272 patients previously untreated with beta-blockers who were also randomly assigned to placebo therapy (group 2). There were no significant differences between the two groups in MB creatine kinase isoenzyme (15.8 +/- 10.9 vs 18.2 +/- 14.4 g-eq/m2, respectively) estimates of infarct size, radionuclide-determined left ventricular ejection fractions within 18 hr of infarction (0.44 +/- 0.15 vs 0.47 +/- 0.16) or 10 days later (0.42 +/- 0.14 vs 0.47 +/- 0.16), creatine kinase-determined incidence of infarct extension (13% vs 6%), congestive heart failure (43% vs 37%), nonfatal ventricular fibrillation (5% vs 7%), or in-hospital mortality (13% vs 9%). Patients in group 1 had more recurrent ischemic chest pain (p = .002) within the first 24 hr after infarction, but not thereafter. However, this did not appear to be related to a rebound increase in systolic blood pressure, heart rate, or double product. In a separate analysis, 20 propranolol-eligible group 1 patients randomly selected for withdrawal of beta-blockade (group 3) were compared with 15 patients randomly selected for continuation of prior beta-blockade therapy (group 4). This comparison yielded similar results. These data indicate that the beta-blockade withdrawal phenomenon is not a major clinical problem in patients with acute myocardial infarction. beta-Blockade therapy can be discontinued abruptly during acute myocardial infarction if clinically indicated.

The time from onset of symptoms to arrival in the hospital emergency room (ER) was studied in 778 patients randomized into a study of acute myocardial infarction (AMI) size limitation. Patients at relatively high risk of death after AMI (including those with preexisting diabetes mellitus, systemic hypertension or congestive heart failure), women and older patients arrived significantly later in the ER than did patients without these characteristics. A significantly higher mortality rate was observed in patients who arrived late, i.e., those who arrived more than 2 hours after the onset of chest pain, even though patients with hemodynamic compromise (bradycardia, hypotension) tended to arrive earlier. The difference in long-term mortality between those who arrived early (within 2 hours of onset of chest pain) and those who arrived late was accounted for by the baseline differences between these 2 groups. These baseline differences may influence the effects of early interventions in AMI. In addition, these findings have implications for education of high-risk patients who could benefit the most from aggressive early intervention.

The results of the propranolol limb of the Multicenter Investigation of the Limitation of Infarct Size are reviewed. A total of 269 patients, who presented with symptoms and electrocardiographic signs suggesting acute myocardial infarction were randomized to acute intravenous and subsequent oral therapy with propranolol (n = 134) or placebo (n = 135). Eligibility for acute beta-blocker therapy was determined on the basis of readily available, noninvasive tests. Therapy was started at an average time of 8.5 hours after onset of symptoms. The full induction dose of intravenous propranolol (0.1 mg/kg) was tolerated by 90% of treated patients, and oral maintenance therapy was being continued in 82% of treated patients on the second hospital day. There was a significant reduction in heart rate throughout maintenance therapy with propranolol, which continued through the tenth hospital day. There was no significant difference in the incidence of congestive heart failure between propranolol- and placebo-treated groups. There was also no significant difference between the 2 groups in infarct size estimated by measurement of serum CK-MB, planimetry of infarct area on technetium pyrophosphate myocardial scintigrams or R-wave measurements in patients with transmural anterior and inferior infarcts. There was no significant difference in mortality between the 2 groups during an average of 36 months' follow-up. Although propranolol can be administered safely to patients with acute myocardial infarction who are selected on the basis of simple clinical criteria, there is no evidence of reduction of infarct size when beta blockade is begun 8.5 hours after the onset of symptoms.

Risk of sudden death was assessed in 533 patients who survived 10 days after acute myocardial infarction (AMI) and were followed for up to 24 months (mean 18) in the Multicenter Investigation of the Limitation of Infarct Size. Analysis of clinical and laboratory variables measured before hospital discharge revealed that the QT interval, either corrected (QTc) or uncorrected for heart rate, did not contribute significantly to prediction of subsequent sudden death or total mortality. In this population, frequent ventricular premature complexes (more than 10 per hour) on ambulatory electrocardiographic monitoring and left ventricular (LV) dysfunction (radionuclide LV ejection fraction of 0.40 or less) identify patients at high risk of sudden death. In patients with these adverse clinical findings, the QTc was 0.468 +/- 0.044 second among those who died suddenly and 0.446 +/- 0.032 second in survivors, and was not statistically significant as an additional predictor of sudden death. Consideration of the use of type I antiarrhythmic agents, digoxin, presence of U waves and correction for intraventricular conduction delay did not alter these findings. Although QT-interval prolongation occurs in some patients after acute myocardial infarction, reduced LV ejection fraction and frequent ventricular premature complexes are the most important factors for predicting subsequent sudden death in this patient population.

Recent observational studies have led to controversy as to whether digoxin therapy following confirmed or suspected myocardial infaction increases mortality. We analyzed the mortality experience of 903 patients enrolled in the Multicenter Investigation of Limitation of Infarct Size (MILIS). Digoxin therapy was not randomly assigned; as in the previous observational studies, the decision to treat or not to treat with digoxin was made by the patient’s personal physician on the basis of the usual clinical indications. Cumulative mortality was 28 % for the 281 digoxin-treated patients versus 11 % for the 622 patients who did not receive digoxin (p < 0.001; follow-up interval from 6 days to 36 months, mean = 25.1 months). However, patients treated with digoxin exhibited more baseline characteristics predictive of mortality than their counterparts. A previous myocardial infarction had occurred in 31 % of the digoxintreated patients versus 12 % of the non-treated patients; prior angina occurred in 50 % versus 32 %, respectively, and the mean ejection fraction was 8.6 ± 1.0 % (SEM) lower for digoxin-treated patients. Adjustment for these difference was performed with three different applications of the Cox method on the basis of: 1) a composite MILIS Risk Score for prediction of mortality in the general MILIS population; 2) a special imbalance risk score composed of 19 variables for which the digoxin and non-digoxin population varied and; 3) a subset of the variables used in the CASS study. These adjustments yielded p values of 0.04, 0.34 and 0.14, respectively, for tests of difference in mortality curves, indicating an absence of significant excess mortality associated with digoxin when appropriate baseline variables were available for adjustment. Thus, findings in the MILIS Database do not support the assertion that digoxin therapy is excessively hazardous post infarction, but the existence of an undetected harmful effect can only be excluded with a randomized study. Until the results of such a study are available, we recommend 1) careful consideration if treatment of left-ventricular dysfunction is actually needed, 2) consideration of alternatives to digoxin therapy, and 3) restriction of digoxin use to the subgroup of patients with persistent CHF and evidence of cardiomegaly.

Ischemic ventricular dysrhythmias were produced in 40 of 47 anesthetized mongrel dogs by high ligation of the left anterior descending coronary artery. Dysrhythmias were treated with a single iv bolus of 20, 40, 80, or 120 mg of lidocaine (L) in order to determine the dose at which approximately 50% of animals had an antidysrhythmic response. Cardiac output and regional myocardial blood flow (RMBF) were measured by using radionuclide labeled microspheres. Lidocaine concentration [( L]) was measured from samples of arterial and venous blood and normal and ischemic myocardium. All dogs treated with 40, 80, or 120 mg of L had an antidysrhythmic effect. However, with 20 mg of L the dysrhythmia persisted in 12 and resolved in 14. With 20 mg of L, ischemic myocardial [L] was greater in dogs with an antidysrhythmic effect than in those with persistent dysrhythmias (1.14 +/- 0.12 vs. 0.76 +/- 0.04 micrograms X g-1), but no difference was seen for arterial, venous, and normal myocardial [L]. Ischemic RMBF was higher in the dogs that had an antidysrhythmic effect than in those that did not, 9.8 +/- 1.5 versus 6.9 +/- 1.3% of normal. With 20 mg of L, [L] in ischemic myocardium correlated well with ischemic RMBF. The antidysrhythmic response to L had a threshold at a tissue concentration of greater than or equal to 1.0 microgram X g-1 (chi-square = 8.55, P less than 0.005). For this model, the [L] in ischemic myocardium during acute ischemia correlates with the antidysrhythmic response to L, while the concentration in normal myocardium or blood does not.

Fourteen patients with transmural acute myocardial infarction (AMI) were treated with intravenous streptokinase a mean of 4 ± 1 hours after chest pain and underwent technetium-99m stannous pyrophosphate (Tc-99m-PPi) imaging 7 ± 2 hours after the onset of chest pain. The early Tc-99m-PPi images were obtained to test the hypothesis that an early, strongly abnormal Tc-99m-PPi image suggests reperfusion. Eleven of 14 patients had early peaking (within 16 hours) serum creatine kinase isoenzyme levels (CK-B) at a mean of 11 ± 3 hours. Ten of 14 patients had 3+ or 4+ acute Tc-99m-PPi images. Eight of 11 patients had patent infarct-related vessels at cardiac catheterization 15 days after AMI. One patient who had both an early positive Tc-99m-PPi image and CK-B peak level had an occluded infarct-related artery at catheterization. Acute left ventricular (LV) ejection fraction (EF) by radionuclide ventriculography was compared with LVEF on day 15, and improved from 0.37 ± 0.13 to 0.50 ± 0.16 (p = 0.004) in the 10 patients with strongly positive acute Tc-99m-PPi images. LVEF also improved from 0.37 ± 0.12 to 0.49 ± 0.15 (p = 0.003) in the 11 patients with early peaking serum CK-B values. Three patients without evidence of reperfusion failed to improve the LVEF from the initial value to the one obtained at hospital discharge. Six control patients had acute Tc-99m-PPi images 10 ± 2 hours after chest pain; none had strongly positive acute Tc-99m-PPi images, and the mean time to peak CK-B was 19 ± 5 hours. Strongly positive early Tc-99m-PPi images may be as reliable as early peaking CK-B in predicting successful reperfusion in patients receiving intravenous streptokinase therapy for AMI.

Methods for detecting acute myocardial infarction (AMI) were compared in a prospective study of 726 patients with pain presumed to be caused by ischemia that lasted 30 minutes or longer and was associated with electrocardiographic changes (ST-segment deviation greater than or equal to 0.1 mV and/or new Q waves or left bundle branch block). Using MB-CK values of more than 12 IU/liter as the standard criterion for detection of AMI, 639 patients (88%) were judged to have AMI. Total plasma CK values, technetium-99m stannous pyrophosphate images 48 to 72 hours after admission, and serial 12-lead electrocardiograms over 10 days were analyzed by investigators blinded to other clinical and laboratory data. For detection of AMI, total CK, electrocardiograms (ECGs) and pyrophosphate imaging were all highly accurate and sensitive (total CK accuracy 97%, ECG 92%, pyrophosphate 88%; total CK sensitivity 98%, ECG 96% and pyrophosphate 91%). However, both pyrophosphate and ECG were less specific than total CK (p less than 0.01) (total CK specificity 89%, pyrophosphate 64% and ECG 59%). The sensitivity (p less than 0.05) and accuracy (p less than 0.01) of total CK and pyrophosphate for those patients with Q-wave development were slightly greater than for those in whom Q waves did not evolve. The ECG was less accurate (p less than 0.02) and pyrophosphate was less specific (p less than 0.04) in patients with prior MI compared with those with initial infarction.(ABSTRACT TRUNCATED AT 250 WORDS)

To characterize the potential of nifedipine in the therapy of unstable angina pectoris we implemented a blinded, randomly assigned, titrated schedule of conventional therapy (propranolol, if not contraindicated, and isosorbide dinitrate) or nifedipine for 14 days in 126 patients hospitalized in a coronary care unit for ischemic chest pain of less than 45 min duration. There were no significant differences between conventionally and nifedipine-treated patients with regard to (1) the time to relief of pain as judged by life table analysis, (2) the decrease in anginal attacks per 24 hr from day 0 to day 2 (-2.5 +/- 0.4 for conventional therapy vs; -2.8 +/- 0.3 for nifedipine), (3) the decrease in the number of nitroglycerin tablets consumed per 24 hr (-2.0 +/- 0.5 for conventional vs -2.1 +/- 0.4 for nifedipine therapy), (4) the percentage of patients requiring morphine on day 1 (13% for conventional vs 21% for nifedipine therapy), or (5) the percentage of patients who developed infarction (14% in both groups). Among the 27 patients who did not respond to initial conventional (n = 13) or nifedipine therapy (n = 14), five in each group became pain free when the opposite therapy (either nifedipine or conventional therapy) was added. In the subgroup of 67 patients who were receiving propranolol before randomization, addition of nifedipine was more effective in controlling pain than was an increase in conventional therapy (p = .026). In the subgroup of 59 patients not receiving prior propranolol, initiation of conventional therapy produced more rapid pain relief than initiation of nifedipine therapy alone (p less than .001), which tended to increase heart rate. Thus, for the study population as a whole therapy with nifedipine alone was equivalent to conventional therapy for unstable angina, although this overall equivalence may result from a combination of superiority of nifedipine therapy in patients previously receiving beta-blocker therapy and superiority of beta-blocker therapy in patients not previously receiving beta-blockers.

To test the hypothesis that single-photon emission computed tomography (SPECT) of 99mTc-pyrophosphate (99mTc-PPi) with and without the overlay of tomographic blood pool scintigrams might detect small infarcts not identified by planar imaging, 52 patients were studied 3.2 +/- 2.0(SD) days after hospital admission for suspected acute myocardial infarction. Patients were chosen prospectively for tomographic study primarily, but not exclusively, because planar four-view imaging with 99mTc-PPi was either negative or equivocal. SPECT was performed with a commercial rotating detector system immediately after planar imaging on one occasion. Corresponding 99mTc-PPi and blood pool sections were mapped into opposite halves of a bichromic color table and displayed as an overlay. Planar images, SPECT and SPECT with blood pool overlay were interpreted separately and in random order without knowledge of clinical data. Seventeen patients had transmural infarcts (four anterior, 13 inferior), 19 had nontransmural infarcts, and 16 patients did not have acute myocardial infarction. The sensitivity of SPECT with blood pool overlay was significantly better than planar imaging for the entire group with myocardial infarction (97% vs 78%; p less than .025); this was primarily due to increased sensitivity in the detection of nontransmural myocardial infarction (95% vs 67%; p less than .05), although in one additional patient inferior transmural myocardial infarction was also detected by the SPECT overlay technique. The specificities of the SPECT overlay technique and planar imaging were not significantly different; however, receiver operating characteristic analysis showed enhanced observer confidence with the tomographic method. SPECT without overlay was intermediate in sensitivity and specificity.(ABSTRACT TRUNCATED AT 250 WORDS)

Preliminary clinical and laboratory observations suggest that nifedipine might prevent progression of threatened myocardial infarction by reversing coronary spasm or might limit necrosis during the course of acute myocardial infarction. We screened 3143 patients with ischemic pain of greater than 45 min duration and randomly assigned 105 eligible patients with threatened myocardial infarction and 66 with acute myocardial infarction to receive nifedipine (20 mg orally every 4 hr for 14 days) or placebo plus standard care. Treatment was started 4.6 +/- 0.1 hr after the onset of pain. Infarct size index was calculated by the MB-creatine kinase (CK) method and expressed as CK-geq/m2 +/- SE. The incidence of progression to infarction among patients with threatened myocardial infarction was not significantly altered by nifedipine (36 of 48 [75%] for placebo-treated and 43 of 57 [75%] for nifedipine-treated patients). Furthermore, infarct size index was similar among placebo- and nifedipine-treated patients (16.9 +/- 1.5 MB-CK-geq/m2, n = 65, and 17.0 +/- 1.5 MB-CK-geq/m2, n = 68, respectively) with threatened myocardial infarction who exhibited infarction and for those with acute myocardial infarction. Among the 171 eligible patients randomly assigned to drug or placebo, 6 month mortality did not differ significantly (8.5% for placebo vs 10.1% for nifedipine, NS), but mortality in the 2 weeks after randomization was significantly higher for nifedipine-treated patients (0% for placebo compared with 7.9% for nifedipine, p = .018).(ABSTRACT TRUNCATED AT 250 WORDS)

One hundred nine patients with persistently positive technetium-99m pyrophosphate (Tc-99m-PPi) myocardial scintigrams 6 months after acute myocardial infarction (MI) (Group A) and 185 patients without such persistently positive scintigrams (Group B) were compared with regard to enzymatically determined infarct size, early and late measurements of left ventricular (LV) function determined by radionuclide ventriculography, and preceding clinical course during the 6 months after MI. The CK-MB-determined infarct size index in Group A (17.4 ± 10.6 g-Eq/m2) did not differ significantly from that in Group B (16.0 ± 14.6 g-Eq/m2). Similarly, myocardial infarct areas in the 2 groups, determined by planimetry of acute Tc-99m-PPi scintigrams in those patients with well-localized 3+ or 4+ anterior pyrophosphate uptake, were not significantly different (35.7 ± 13.4 vs 34.4 ± 13.1 cm2, respectively). However, patients in Group A had significantly lower LV ejection fractions than those in Group B, both within 18 hours of the onset of MI (0.42 ± 0.14 vs 0.49 ± 0.14, p < 0.01) and at 3 months after MI, both at rest (0.42 ± 0.14 vs 0.51 ± 0.14, p < 0.01) and at maximal symptom-limited supine bicycle exercise (0.44 ± 0.17 vs 0.51 ± 0.17, p < 0.01). Peak exercise levels achieved in the 2 groups were not significantly different. Furthermore, patients in Group A demonstrated a greater incidence of congestive heart failure during the initial hospital admission (41 vs 24%; p < 0.01) and a greater requirement for digoxin (p < 0.05) and furosemide (p <0.01) after discharge. These results were not explained by an y difference in the incidence of previous MI between the 2 groups.

The influence of basal heart rate (HR) on the effects of inotropic (dobutamine infusion) and chronotropic (atrial pacing) stimuli during acute myocardial ischemia was assessed by measurement of intramural carbon dioxide partial pressure (PCO2) in open-chest dogs undergoing transient 10-minute left anterior descending coronary artery occlusions. In protocol I, in 5 dogs anesthetized with pentobarbital alone, HR increased from 153 ± 10 to 182 ± 7 beats/min between experimental coronary occlusions, but the increase in ischemic zone intramural carbon dioxide tension (ΔPmCO2) was not altered by this significant increase in HR. In protocols II to V, dogs were anesthetized with combinations of morphine, thiamylal and pentobarbital and had a basal average HR of 81 beats/min. Atrial pacing in protocol II (13 dogs) increased HR from 76 ± 21 to 134 ± 19 beats/min (p < 0.001); left ventricular (LV) myocardial oxygen consumption (MVO2) rose from 3.9 ± 1.6 to 4.9 ± 1.4 ml/min/100 g (p < 0.05), and (ΔPmCO2) rose from 42 ± 14 to 50 ± 15 mm Hg (p < 0.01), indicating more severe ischemic injury during the second experimental coronary occlusion. In protocol III, 11 dogs received 20 μg/kg/min of dobutamine before the second experimental occlusion, which significantly (p < 0.02) increased HR, LV dPdt and MVO2; (ΔPmCO2) increased from 46 ±13 to 63 ± 18 mm Hg (p < 0.01). The 7 dogs in protocol IV received 3.9 ± 1.9 μg/kg/min of dobutamine, titrated such that HR was unchanged (84 ±10 vs 81 ± 15 beats/min). but LV dPdt increased by 92% (p < 0.001). The (ΔPmCO2) was unchanged between experimental occlusions in this group. In protocol V, 9 dogs had atrial pacing before both experimental coronary occlusions (HR 133 ± 7 and 174 ± 7 beats/min, respectively, p < 0.001). The (ΔPmCO2) was similar during both periods of ischemia. In 4 control dogs anesthetized with morphine, thiamylal and pentobarbital, there was no significant change in (ΔPmCO2) during 3 successive coronary occlusions performed without interventions.

Enzymatic estimates of myocardial infarct size based on plasma levels of MB creatine kinase (MB-CK) were compared with anatomic infarct size in 49 human hearts obtained at autopsy. The patients studied had been enrolled in the Multicenter Investigation of Limitation of Infarct Size (MILIS) study program within 18 hr of the onset of acute infarction and were treated at one of five participating hospitals. Infarct size was estimated from serial measurements of plasma MB-CK made at the core laboratory for CK analysis. Hearts obtained at autopsy were studied independently by the core pathology laboratory without knowledge of the MB-CK levels or clinical results. Data from the two laboratories were compared at the data coordinating center. Of 49 hearts, 12 were excluded either because anatomic infarct size could not be established or because the infarct occurring at the time of enrollment in the MILIS study could not be distinguished with certainty from other infarcts. Of the remaining 37 hearts, peak MB-CK level was available in 36, but samples sufficient for estimation of infarct size were available in only 25. The overall correlation coefficient (Spearman) was .87 for these 25 hearts, indicating that enzymatic estimates of infarct size correlate closely with anatomic measurements. The results indicate that CK estimates of myocardial infarct size represent a valid clinical end point for assessing myocardial infarct size, and the effect of therapy thereon, in groups of treated and control patients.