We previously found that superoxide dismutase (SOD) did not limit myocardial infarct size after 40 or 90 minutes of ischemia and 4 days of reperfusion in dogs. Because some other studies have shown limitation of infarct size after shorter periods of reperfusion, we postulated that our negative results might be due to late reperfusion injury mediated by superoxide anions produced after excretion of SOD. To test this "early protection-delayed death" hypothesis, we have examined whether SOD, conjugated to polyethylene glycol (PEG-SOD) to prolong its circulating half-life, limited myocardial infarct size. The circumflex artery was occluded for 90 minutes followed by 4 days of reperfusion. PEG-SOD (total dose, 10,000 units/kg) and catalase (55,000 units/kg) were given during the 30 minutes before reperfusion. Plasma SOD levels in the treated group were 330 +/- 20 units/ml at the onset of reperfusion and 140 +/- 10 units/ml on day 4 (circulating half-life, 75 +/- 5 hours) versus 5 +/- 1 units/ml in controls. Histological infarct size was 37.1 +/- 4.2% of the area at risk in the treated group (n = 11) versus 44.5 +/- 6.2% in controls (n = 10) (p = NS). Infarct size and collateral blood flow were inversely related in controls; PEG-SOD and catalase did not shift this regression (p = NS by analysis of covariance). Thus, infarct size was not limited when measured after 4 days of reperfusion, even though plasma SOD exceeded 100 units/ml throughout this reperfusion period.(ABSTRACT TRUNCATED AT 250 WORDS)