Clinical practice and laboratory studies have demonstrated the efficacy of cold crystalloid cardioplegia for donor heart protection. Efforts to increase the margin of safety for protection led us to compare unmodified University of Wisconsin (UW) solution to the dextrose, mannitol-based Stanford (ST) solution. A canine model of heart transplantation with antegrade hypothermic cardioplegic arrest and 6 hours of 4 degrees C ischemic storage was used. An oxygenated blood-primed isolated heart preparation was used for reperfusion and myocardial mechanics and energetics studies of the working heart. Six of 6 UW and 4 of 6 ST hearts reached the working phase. Computer-assisted analysis of pressure-volume loops generated at varying flows measured by tri-axial sonomicrometry and high-fidelity micromanometry showed no significant differences in function between the ST and UW groups by maximum elastance (UW, 4.2 +/- 1.1; ST, 4.0 +/- 0.7), preload recruitable stroke work (UW, 43.7 +/- 7.3; ST, 43.4 +/- 8.7), or slope of log-linear end-diastolic pressure-volume curve (UW, 0.057 +/- 0.01, ST, 0.061 +/- 0.01). Specimens for determination of myocardial water content were taken after cardioplegic arrest, after storage, after reperfusion, and after the working phase. There was a significant increase in tissue water after reperfusion in both groups (UW, 75.7% +/- 0.5% to 81.6% +/- 0.2%, p = 0.0001; ST, 76.5% +/- 0.4% to 83.4% +/- 0.3%, p = 0.0002), which persisted after the working phase (UW, 81.5% +/- 0.9%, p = 0.0002; ST, 82.6% +/- 0.1%, p = 0.0003). Both groups exhibited postreperfusion increase in myocardial water content, but this edema was significantly less marked in the UW group (p = 0.0001).(ABSTRACT TRUNCATED AT 250 WORDS)