Intracoronary ultrasound imaging: correlation of plaque morphology with angiography, clinical syndrome and procedural results in patients undergoing coronary angioplasty.

OBJECTIVES: This study was designed to establish the relation between ultrasound-derived atheroma morphology and the clinical, procedural and angiographic features of patients presenting for coronary angioplasty. BACKGROUND: Intracoronary ultrasound imaging provides accurate dimensional information regarding arterial lumen and wall structures. Atheroma composition may also be assessed by ultrasound; however, only limited studies have been performed in patients. METHODS: In 65 patients a diagnostic ultrasound imaging catheter or a combination imaging-angioplasty balloon catheter was used during coronary angioplasty to image both the lesion and the vessel segment just proximal to it (reference segment). Ultrasound images were analyzed for lumen, total vessel and plaque areas and were classified into five morphologic subtypes (soft, fibrous, calcific, mixed plaque and concentric subintimal thickening). These data were compared with angiographic morphologic features, procedural results and clinical angina pattern (stable vs. unstable). RESULTS: Morphologic analysis of the ultrasound images obtained from the lesion correlated well with the clinical angina syndrome. Compared with patients with stable angina, patients with unstable angina had more soft lesions (74% vs. 41%), fewer calcified and mixed plaques (fibrotic, soft or calcific components in one or more combinations [25% vs. 59%]) and fewer intralesional calcium deposits (16% vs. 45%) (all p < 0.01). There was no correlation between ultrasound and angiographic lesion morphologic characteristics for either the reference segment or the lesion. Ultrasound demonstrated greater sensitivity than angiography for identifying unstable lesions (74% vs. 40%). Dimensional analysis demonstrated a large plaque burden in the reference segments (45 +/- 15% of total vessel area). Postangioplasty plaque burden was also high (62 +/- 9%). There was a significant, but only fair correlation between lumen area determined by angiography and ultrasound for both the reference segment (r = 0.70, p < 0.001) and the postangioplasty lesion (r = 0.63, p < 0.05). CONCLUSIONS: Morphologic plaque classification by ultrasound is closely correlated to clinical angina but has little relation to established angiographic morphologic characteristics. Intracoronary ultrasound imaging during angioplasty identifies a large residual plaque burden in both the reference segment and the lesion. In the future, determination of plaque composition by intracoronary ultrasound may be important in selecting or modifying interventional therapeutic options.

Increased left ventricular dysfunction in elderly patients despite successful thrombolysis: the GUSTO-I angiographic experience.

OBJECTIVE: This study sought to determine whether the recovery of regional and global left ventricular function is reduced in elderly patients despite successful thrombolytic therapy for acute myocardial infarction. Comparisons were made between elderly (> or = 75 years old, n = 47) and adult (< 75 years old, n = 434) patients enrolled in the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO) angiographic trial who underwent catheterization at 90 min and 5 to 7 days after thrombolysis and who had an open infarct-related artery with Thrombolysis in Myocardial Infarction (TIMI) grade 2 to 3 flow at both times. BACKGROUND: The morbidity and mortality of acute myocardial infarction is increased in elderly patients, presumably because of multiple adverse coexistent baseline variables. However, functional recovery after thrombolysis has not been characterized in the elderly. METHODS: Ejection fraction, end-systolic volume index, infarct and noninfarct zone contractile function (SD/chord) and infarct extent (number of chords) were determined. RESULTS: At 90 min, elderly patients with an open infarct-related artery had decreased infarct zone contractile function (-2.8 +/- 0.2 vs. -2.3 +/- 0.1 SD/chord in adults, p < or = 0.05) and a greater extent of injury (26.0 +/- 2.6 vs. 20.7 +/- 0.8 chords in adults, p < or = 0.05). At 5- to 7-day follow-up ventriculography, ejection fraction was reduced, and end-systolic volume index was significantly increased in elderly patients compared with adults. The severity of regional wall motion dysfunction in the infarct zone was also greater in the elderly than in adults at 5- to 7-day follow-up (-2.6 +/- 0.2 vs. -1.9 +/- 0.1 SD/chord, respectively, p < or = 0.005). Non-infarct zone contractile function at 90-min ventriculography was similar in both groups. Despite a patent infarct-related artery at 90-min, the 30-day mortality rate in the elderly remained elevated (17.8%) compared with that of adults (4%) (p < or = 0.0001). Elderly patients were predominantly female and had a higher prevalence of hypertension, multivessel coronary disease, previous infarction, anterior infarctions and later time to treatment (between 3 and 6 h) than adults. However, age > or = 75 years remained an independent determinant by multivariable regression analysis of 1-week postinfarction end-systolic volume index, regional left ventricular dysfunction (p = 0.02 and p < or = 0.008, respectively) and 30-day mortality (p < or = 0.0001). CONCLUSIONS: Elderly patients had increased damage in the infarct zone and had persistently increased mortality despite sustained infarct-related artery patency after successful thrombolysis. Although the causes are probably multifactorial, a more rapid progression of ischemic injury or a blunted postreperfusion recovery appears to contribute to the poorer outcomes in elderly patients.

Reduction of infarct size by cell-permeable oxygen metabolite scavengers.

The timely restoration of blood flow to severely ischemic myocardium limits myocardial infarct size. However, experimental studies demonstrate that the myocardial salvage achieved is suboptimal because of additional injury that occurs during reperfusion, due in part to the generation of reactive oxygen metabolites. Initially, superoxide (O2-) was considered to be the central mediator of reperfusion injury. While there are several potential pathways of O2- generation in reperfused myocardium, O2- is poorly reactive toward tissue biomolecules. However, O2-, in the presence of redox-active metals such as iron, generates .OH or hydroxyl-like species that are highly reactive with cell constituents. Thus, while O2- may initiate reaction sequences leading to myocardial injury, it may not be the actual injurious agent. In vitro studies suggest that oxygen metabolite injury occurs at intracellular sites and involves iron-catalyzed processes. Consistent with this mechanism, extracellular oxygen metabolite scavengers have not convincingly reduced infarct size. However, treatment around the time of reperfusion, after ischemia is well established, with cell-permeable scavengers of .OH reduce infarct size. Results with these cell-permeable agents suggest that in the intact animal during regional ischemia and reperfusion, oxygen metabolite injury also occurs at intracellular sites. Cell-permeable scavenger agents are a promising class of drugs for potential clinical use, though further experimental and toxicologic studies are required.

Oxidation and release of glutathione from myocardium during early reperfusion.

Glutathione (GSH) is an important intracellular defense against reactive oxygen metabolites. Reaction of GSH with peroxides generates oxidized glutathione (GSSG). We hypothesized that reperfusion would cause oxidation of GSH and release of GSSG as a potential marker of intracellular oxidative reactions. Ten dogs underwent 90 min left anterior descending (LAD) occlusion and 30 min reperfusion. Coronary sinus (CS) plasma was sampled from the great cardiac vein, which drains the LAD region, and from the aorta at pre-ischemia (I), 90 min ischemia, and during reperfusion (R). We found that both GSSG and GSH increased in coronary sinus plasma during early reperfusion. (Formula: see text) Measured GSSG did not arise from autoxidation of plasma GSH. GSH and GSSG release from myocardium not only may be evidence of intracellular oxidative injury, but loss of GSH also could impair metabolism of peroxides during early reperfusion and predispose to further injury.

Trolox C, a lipid-soluble membrane protective agent, attenuates myocardial injury from ischemia and reperfusion.

The lipophilic antioxidant Trolox C, a vitamin E analog, was administered to isolated, buffer-perfused rabbit hearts subjected to 25 min of global stop-flow ischemia and 30 min of reperfusion. In six hearts, Trolox C (200 microM) was infused for 15 min immediately prior to ischemia and for the first 15 min of reperfusion. Six control hearts received only vehicle. Gas chromatography analysis confirmed that effective myocardial levels of Trolox were attained. At 30 min reperfusion, the recovery of left ventricular developed pressure was 56 +/- 3% of baseline in control hearts versus 70 +/- 4% in Trolox-treated hearts (p < .01). There was also significant improvement in recovery of Trolox-treated hearts in diastolic pressure and both maximum and minimum values of the first derivative of left ventricular pressure (dP/dt). Creatine phosphokinase release into the coronary effluent at 30 min of reperfusion was 16.5 +/- 8.4 IU/min in untreated and 6.3 +/- 1.0 IU/min (p < .05) in Trolox-treated hearts. Thus Trolox C, a lipophilic antioxidant, attenuated myocardial injury during stop-flow ischemia and reperfusion.

Sensitivity of protein sulfhydryl repair enzymes to oxidative stress.

According to their demonstrated activities, the thiol-disulfide oxidoreductase (TDOR) enzyme systems [thioltransferase (glutaredoxin) and GSSG reductase; and thioredoxin and thioredoxin reductase] are expected to provide the primary cellular mechanism for protection and repair of sulfhydryl proteins under oxidative stress. Since all four enzymes have active site dithiol moieties, they may be vulnerable to oxidative damage themselves. Therefore, an hydroxyl radical generating system (chelated ferrous iron in combination with hydrogen peroxide) was used to document the relative sensitivity of each of the enzymes to oxidative stress in vitro. At particular concentrations of enzymes and oxidant system, all of the enzymes were deactivated nearly completely, but different patterns of susceptibility were observed. At the approximate physiological concentration of each enzyme thioredoxin and thiol-transferase were largely deactivated with 1 mM Fe2+-ADP, 1 mM H2O2; whereas thioredoxin reductase and GSSG reductase were much less sensitive: 10 microM thioredoxin (88% deactivated), 1 microM thioltransferase (72%), 2 microM thioredoxin reductase (5%), and 0.1 microM GSSG reductase (17%). As the concentration of the oxidant system was decreased stepwise from 1 mM to 1 microM to mimic conditions that may be associated with oxidative tissue injury in situ, deactivation of thioredoxin was decreased proportionately, whereas thioltransferase remained much more susceptible. As expected GSH and other radical scavengers protected thioltransferase from deactivation by Fe(ADP)-H2O2. To test the susceptibility of the TDOR enzymes to oxidative stress in a physiological-like setting, isolated perfused rabbit hearts were subjected to 30 min ischemia and 30 min reperfusion. The GSH/GSSG ratio and total dethiolase activity (thioltransferase and thioredoxin systems) remained unchanged relative to control hearts, indicating that overall redox status and sulfhydryl repair activity are maintained during moderate oxidative stress in situ.

Hydrogen peroxide decreases effective refractory period in the isolated heart.

Although previous investigations have concluded that reactive oxygen metabolites contribute to reperfusion arrhythmias, the experimental models employed also had a significant amount of tissue injury, which may have contributed to the observed electrophysiologic effects. We studied whether exposure of the intact heart to a reactive oxygen metabolite at doses that are not associated with histologic evidence of cell necrosis would alter myocardial refractoriness, suggesting that subtle and reversible oxidative stress could alter myocardial electrophysiologic properties and perhaps contribute to ventricular arrhythmias. Isolated rabbit hearts were perfused for 30 min with low doses of hydrogen peroxide (H2O2), either 10(-5), 5 x 10(-6), or 10(-6)-M H2O2 versus vehicle alone; followed by a 30-min washout period without H2O2. Infusion of H2O2 for 30 min decreased ventricular epicardial effective refractory period (ERP) in a dose-dependent manner compared to saline controls (delta ERP). The delta ERP versus time curves during the last 10 min of H2O2 infusion were different (p less than 0.01) for each of the three H2O2 doses. Creatine phosphokinase and reversible oxidized glutathione release occurred during 10(-5)-M H2O2 infusion, but not with lower H2O2 doses. Exposure of the intact heart to low concentrations of H2O2, in a range that caused subtle oxidative injury, decreased ventricular ERP in a dose-dependent manner. Thus, H2O2 generation could contribute to ventricular arrhythmias, even in settings of sublethal and potentially reversible oxidative injury.

Dimethylthiourea, but not dimethylsulfoxide, reduces canine myocardial infarct size.

We studied the effect of treatment with two diffusible, low molecular weight scavengers of toxic oxygen metabolites, dimethylthiourea (DMTU) and dimethylsulfoxide (DMSO), on canine infarcts caused by 90 min of ischemia and 3 h of reperfusion. Infarct size was determined by incubating ventricular slices with triphenyl tetrazolium chloride. Areas at risk were determined by autoradiography of 99Tc microspheres injected in vivo during ischemia and were similar (p greater than 0.05) in DMTU, DMSO, and saline treated dogs. However, the ratio of infarct size to area at risk was reduced (p less than 0.05) in dogs treated 30 min before reperfusion with 500 mg/kg DMTU (31.1 +/- 4.6%, n = 9) compared with saline treated dogs (53.4 +/- 4.6% n = 9). In contrast, the ratio of infarct size to area at risk was not significantly different (p greater than 0.05) in dogs treated with 2000 mg/kg DMSO 30 min before reperfusion (43.7 +/- 4.3%) compared to saline treated dogs. The serum concentration of DMTU (4.5 mM) was one-tenth that of DMSO (48 mM) in early reperfusion. Therefore, DMTU but not DMSO protected against post-ischemic cardiac reperfusion injury.

Coronary endothelial dysfunction from ischemia and reperfusion: effect of reactive oxygen metabolite scavengers.

Using anesthetized mongrel dogs exposed to 60 min of ligation of the left anterior descending coronary artery followed by 60 min of reperfusion, we examined the effect of superoxide dismutase (SOD) and dimethylthiourea (DMTU) on evidence of endothelial injury in coronary rings studied in vitro. In 13 dogs treated with saline rings from the normal left circumflex coronary artery (LCF) relaxed by 98 +/- 4% when exposed to 10(-5) M acetylcholine whereas rings from the left anterior descending coronary artery (LAD) relaxed by 79 +/- 7% (p less than 0.05). In the same rings maximum relaxation with the ionophore A23187 was 107 +/- 5% versus 87 +/- 8% (p less than 0.05) for the LCF and the LAD, respectively. Comparisons of concentration-response curves through a range of doses of both acetylcholine and A23187 revealed significant differences for both vasodilators between the LCF and the LAD (p less than 0.01 for each). Nine dogs were treated with bovine SOD infused in the left atrium the last 20 min of ligation and throughout reperfusion (140 units/kg/min) and six other dogs were treated with DMTU 500 mg/kg i.v. given the last 30 min of the ligation period. Neither SOD nor DMTU prevented endothelial injury in the LAD. Despite pretreatment with these agents, there were significant reductions in maximum relaxation and in total concentration-response curves in the LAD as compared with the results in rings from the LCF with both acetylcholine and A23187. There were normal responses to nitroprusside in both the LCF and LAD in all three experimental groups.(ABSTRACT TRUNCATED AT 250 WORDS)

ATP catabolism and adenosine generation during ischemia in the aging heart.

Myocardial injury following ischemia and reperfusion is increased in the aging heart. The mechanisms underlying the increased susceptibility of the aging heart to ischemic injury remain unknown. We investigated whether decreased glycogen utilization with a more rapid depletion of ATP occurred during ischemia in the aging heart. Isolated buffer-perfused hearts from adult (6 months old) and aging (24 months old) Fischer 344 rats were subjected to 0, 2, 5, 10, 15 or 25 min of global stop-flow ischemia following a 15 min equilibration period (n = 5-6 for each ischemic time at each age). ATP level were decreased at preischemic baseline in aging hearts. ATP levels remained lower in the aging heart throughout ischemia (P < 0.001) with a similar pattern of decrease in both age groups. The decrease in tissue glycogen and increase in lactate contents was similar during ischemia in both age groups, suggesting that comparable glycogen utilization occurred during ischemia in adult and aging hearts. ATP catabolism leads to ADP, AMP and then adenosine. Tissue levels of adenosine, an important cardioprotective metabolite, were measured during ischemia. Tissue adenosine levels were decreased by 50% in the aging heart at 5 and 10 min, and remained depressed at 15 min and 25 min of ischemia compared to adult controls. Thus, increased ischemic injury in the aging heart is not related to differences in glycogen consumption. Lower tissue ATP levels and decreased adenosine levels were observed during ischemia. The differences in ATP content between adult and aging hearts occurred only during early ischemia and are unlikely to provide a mechanism for the increased damage observed following more prolonged periods of ischemia in the aging heart. The potential contribution of these decreases in tissue adenosine content to the increased injury observed in the aging heart will require further study.

Effects of acute left anterior descending occlusion on regional myocardial blood flow and wall thickening in the presence of a circumflex stenosis in dogs.

In this study, 2 hypotheses were tested: (1) Myocardium supplied by a stenosed circumflex coronary artery (LC) does not demonstrate compensatory increases in regional blood flow and systolic thickening when the left anterior descending coronary artery (LAD) is acutely occluded. (2) Blood flow to myocardium in the distribution of an acutely occluded LAD is lower in the presence of a stenosed than in the presence of an unstenosed LC. Fifty-three open-chest, anesthetized dogs were studied. Regional coronary blood flow (8 to 10-mu microspheres) and wall thickening (sonomicrometer crystals) were measured before and after LAD occlusion in the presence of an unstenosed LC artery, and a moderate and severe LC stenosis. Acute LAD occlusion in the presence of an unstenosed LAD was not accompanied by a significant increase in regional blood flow to the remote LC bed; posterior myocardial wall thickening, however, increased from 0.22 +/- 0.02% to 0.24 +/- 0.02% (p = 0.04). In the presence of a moderate LC stenosis (gradient 29 +/- 1 mm Hg), LAD occlusion was associated with a 9% (p = 0.02) decrease in endocardial flow and an 11% decrease in the endocardial/epicardial flow ratio (p = 0.002). Transmural flow was unchanged and there was no compensatory increase in posterior wall thickening. In the presence of a more severe LC stenosis (gradient 49 +/- 1 mm Hg), central LC endocardial flow decreased by 32% (p = 0.0008) at the time of LAD occlusion. Similar alterations were noted in the peripheral LC region.(ABSTRACT TRUNCATED AT 250 WORDS)

Tissue iron overload and mechanisms of iron-catalyzed oxidative injury.

Tissue iron overload causes clinical syndromes that involve the heart, liver, and pancreas. While tissue iron uptake occurs by both transferrin-dependent and independent processes, tissue uptake in the iron overload syndromes occurs predominantly via transferrin-independent mechanisms. Increased redox-active iron present in hemeproteins and the cytosolic iron pool can catalyze oxidative damage to lipids, proteins, and nucleic acids, either by oxyradical dependent or independent mechanisms. Iron-catalyzed injury results in damage to cell constituents, including mitochondria, lysosomes, and the sarcolemmal membrane. These mechanisms of iron-mediated damage are involved in the pathogenesis of organ dysfunction in primary hemochromatosis, transfusion-related iron overload, ischemia-reperfusion injury, and cardiac anthracycline toxicity.

Incremental iron overload during reperfusion progressively augments oxidative injury.

OBJECTIVE: To determine if a relationship exists between the extent of iron-catalyzed injury and the degree of tissue iron overload during reperfusion. METHODS: To selectively increase tissue iron only during early reperfusion, isolated, buffer perfused rabbit hearts were exposed to 20 microM Fe(2+)-100 microM ADP during the last 3 minutes of ischemia and the initial 4 minutes of reperfusion. Control groups were exposed to ADP and iron-ADP regimens that did not increase intracellular iron. All the hearts received 30 minutes of normothermic global ischemia and 30 minutes of reperfusion. Heart function was monitored continuously throughout each experiment. Tissue iron and biochemical markers were analyzed at the end of experiments. RESULTS: Hemodynamic recovery was decreased and tissue lipid peroxide levels were increased in the 20 microM Fe(2+)-100 microM ADP group compared to controls. The recoveries of developed pressure and positive/negative dP/dT at 30 minutes of reperfusion were negatively correlated with tissue iron levels, while cytosol and membrane lipid peroxide levels correlated positively with the iron levels during reperfusion. CONCLUSION: The extent of oxidative injury during reperfusion was directly related to the tissue iron burden present during reperfusion. Increased lipid peroxidation was the principal chemical marker of iron-catalyzed injury.

Exogenous intracellular, but not extracellular, iron augments myocardial reperfusion injury.

Although previous studies using iron chelators suggest that iron-catalyzed reactions exacerbate myocardial injury, a direct demonstration of the timing, sites, and mechanisms of iron-mediated damage during reperfusion has been lacking. Catalytic doses of redox-active iron react with exogenously administered oxygen radical-generating systems to exacerbate myocardial injury. In an analogous manner, catalytic doses (5 microM) of excess iron present during early reperfusion should augment oxidative injury, if the redox-active iron is present in the same compartment as both the oxygen radicals generated during reperfusion as well as the critical biochemical targets of oxidative injury. We determined whether catalytic doses of iron given during early reperfusion could exacerbate myocardial injury and whether iron-catalyzed injury required intra- or extracellular iron. Buffer-perfused rabbit hearts underwent 30 min of 37 degrees C global ischemia and 30 min of reperfusion. Iron (5 microM), attached to ligands that either restrict iron to the extracellular space (ADP) or facilitate the entry of iron into myocytes (omadine, tropolone), was infused during the last 3 min of ischemia and the first 4 min of reperfusion. Recovery of developed pressure was decreased (P < 0.05) in omadine-iron and tropolone-iron compared with ADP-iron and noniron hearts treated with ligands alone. Tissue lipid peroxide levels, an index of oxidative injury, were increased (P < 0.05) by omadine-iron and tropolone-iron, but not ADP-iron. The oxidative damage caused by omadine-iron was blocked by pretreatment with dimethylthiourea, a cell-permeable scavenger of the hydroxyl radical.(ABSTRACT TRUNCATED AT 250 WORDS)

Myocardial sulfhydryl pool alterations occur during reperfusion after brief and prolonged myocardial ischemia in vivo.

Myocardial sulfhydryl (SH)-containing compounds, including reduced glutathione (GSH), are both defenses against and potential markers of reactive oxygen metabolite injury during ischemia and reperfusion. We examined the alterations in GSH and other myocardial SH pools during reperfusion in anesthetized dogs exposed to brief (15 minutes, n = 7) or prolonged (90 minutes, n = 6) regional ischemia caused by occlusion of the left anterior descending artery. Ninety minutes of ischemia followed by 5 hours of reperfusion, which resulted in myocardial necrosis of 43.9 +/- 4.0% of the area at risk, caused a 22% reduction in total myocardial SH groups (p less than 0.01), a 57% decrease in nonprotein myocardial SH groups (p less than 0.01), a 56% decrease in GSH (p less than 0.01), and a 62% decrease in non-GSH, nonprotein SH groups (p less than 0.02). However, protein SH groups were not significantly reduced (12% decrease, p = NS). Also, myocardial release of GSH and oxidized glutathione (GSSG) into the coronary venous effluent occurred during early reperfusion. In contrast, 15 minutes of ischemia, followed by 30 minutes of reperfusion, did not alter myocardial total SH groups, protein SH groups, or GSH (9% decrease, p = NS); nor was there reperfusion release of GSH or GSSG. However, even with brief ischemia, nonprotein SH groups decreased 23% (p less than 0.05), due mainly to a 59% decrease in the non-GSH, nonprotein SH pool (p less than 0.05). These changes after brief ischemia occurred without alterations in myocardial GSSG or the GSH/GSSG ratio.(ABSTRACT TRUNCATED AT 250 WORDS)

Safety of cardiac catheterization via peripheral vascular grafts.

There are few data concerning the complications and technical difficulties encountered when cardiac catheterization is performed using peripheral bypass grafts for vascular access. All cardiac catheterizations performed at our institution from January 1, 1984 to April 1, 1991 were retrospectively reviewed to assess the in-hospital clinical outcomes in patients who had arterial access for catheterization achieved via prosthetic graft puncture. Seventeen procedures had percutaneous puncture of a vascular graft from a total of 2,929 arterial catheterizations performed. The interval from graft placement to catheterization was 7.5 +/- 1.1 years. Arterial sheaths were employed in all cases and corresponded to the catheter size, with 5F systems used in 53% and 7F or larger systems used in the remaining patients. No intraprocedural or postprocedural complications were recognized. Technical difficulties were limited to the inability to selectively cannulate a nondominant right coronary artery in 1 patient. We conclude that percutaneous introduction of an arterial sheath and left heart catheterization via remotely implanted vascular bypass grafts is not associated with an increased risk of procedural complications or technical difficulties.

Deferoxamine pretreatment reduces canine infarct size and oxidative injury.

To test whether iron-catalyzed processes contribute to myocardial necrosis during ischemia and reperfusion, we administered the iron chelator, deferoxamine, to chloralose-anesthetized dogs subjected to 90 min of left anterior descending artery occlusion followed by 360 min of reperfusion. Deferoxamine blocks iron-catalyzed hydroxyl radical formation in vitro. Groups of dogs received either pretreatment with deferoxamine or iron-loaded deferoxamine (15 mg/kg over 30 min preocclusion and 2.5 mg/kg/hr during the first 120 min of reperfusion), equal volumes of saline or deferoxamine treatment during reperfusion (15 mg/kg over 30 min beginning at 75 min of occlusion followed by 2.5 mg/kg/hr during the remainder of the first 120 min of reperfusion). Infarct size as a percentage of area at risk was reduced (P less than .05) by deferoxamine pretreatment (29.8 +/- 4.8%, n = 7, +/- S.E.) compared to saline control (46.8 +/- 4.7%, n = 8), deferoxamine reperfusion (50.5 +/- 6.7%, n = 8) or iron-loaded deferoxamine (60.2 +/- 8.6%, n = 3)-treated dogs. Deferoxamine pretreatment also decreased (P less than .05) the release of oxidized glutathione into the coronary sinus during early reperfusion compared to the other groups. There were no differences between groups in area at risk, risk zone blood flow during ischemia or in heart rate-blood pressure product. Deferoxamine did not decrease hydrogen peroxide concentration, neutrophil superoxide anion production or neutrophil adherence in vitro. We conclude that iron-mediated processes, possibly including iron-catalyzed hydroxyl radical formation, contribute to myocardial necrosis during regional ischemia and reperfusion.