Determination of left ventricular mass by real-time three-dimensional echocardiography: in vitro validation.

Twenty-one explanted fixed hearts (14 dogs and 7 pigs) were examined to validate newly developed real-time three-dimensional (RT3D) echocardiography for measurement of left ventricular (LV) mass in vitro and to compare its accuracy and variability with those of conventional echocardiographic measurements. There was an excellent correlation and high degree of agreement for the determination of LV mass between RT3D echocardiography and true mass measurement (r = 0.98; standard error of the estimate [SEE] = 7.3 g; absolute difference [AD] = 2.8 g; y = 1.00 x -4.0, interobserver variability; 5.0%). The conventional echocardiographic methods yielded weaker correlations, larger standard errors, and interobserver variability (area-length method: r = 0.90; SEE = 13.3 g; AD = 13.2 g; 13.3 % / truncated ellipsoid method: r = 0.91; SEE = 14.7 g; AD = 10.5 g; 7. 9% / M-mode: r = 0.91; SEE = 16.2 g; AD = 9.4 g; 15.3%). Determination of LV mass by RT3D echocardiography has a high degree of accuracy and is superior to conventional one- and two-dimensional echocardiographic methods.

Improvement of rejection-induced diastolic abnormalities in rat cardiac allografts with inducible nitric oxide synthase inhibition.

OBJECTIVE: Inhibition of inducible nitric oxide synthase (nitric oxide II) activity has been proposed as a method to attenuate capillary leak and edema during rejection of heterotopically transplanted rat hearts. Myocardial edema has previously been implicated in diastolic dysfunction during allograft rejection. Accordingly, we tested the hypothesis that inducible nitric oxide synthase inhibition with aminoguanidine would alleviate left ventricular stiffening and myocardial edema formation in 4-day heterotopic rat heart allografts. METHODS: Passive left ventricular filling was studied in American Cancer Institute Lewis rats receiving heterotopic heart transplants receiving either aminoguanidine, a selective nitric oxide synthase inhibitor (n = 6); dexamethasone (1 mg. kg(-1). d(-1) administered subcutaneously) for 4 days after transplantation (n = 6); or intravenous saline solution (n = 6). American Cancer Institute-to-American Cancer Institute isografts (n = 6) were used as controls. RESULTS: Serum nitrite/nitrate levels in the aminoguanidine group (18 +/- 3 mmol/L) and dexamethasone group (22 +/- 4 mmol/L) were reduced versus the intravenous saline group (144 +/- 36 mmol/L [SEM]) to levels seen in controls (25 +/- 9 mmol/L). Left ventricular volume at 15 mm Hg for the aminoguanidine group was increased versus that for the intravenous saline solution group, similar to that for controls, and reduced versus dexamethasone-treated animals. Myocardial water content for the aminoguanidine-treated animals (78.3% +/- 0.4%) was similar to those of intravenous saline-treated animals (78.0% +/- 0. 3%) but greater than those of controls (77.1% +/- 0.2%) and dexamethasone-treated animals (76.7% +/- 0.3%). CONCLUSIONS: Nitric oxide II inhibition with aminoguanidine minimizes the reduction in left ventricular filling that is seen with allograft rejection through a mechanism that is not associated with attenuation of myocardial edema.

Risk stratification for coronary bypass surgery in patients with left ventricular dysfunction: analysis of the coronary artery bypass grafting patch trial database.

BACKGROUND: Preoperative characteristics may influence morbidity and mortality in patients undergoing coronary artery bypass grafting (CABG). The CABG Patch Trial was designed to assess the impact of prophylactic insertion of an implantable cardioverter-defibrillator in patients undergoing high-risk CABG. This database was used to investigate the influence of symptomatic congestive heart failure (CHF) and angina on morbidity and mortality in CABG patients with ventricular dysfunction. METHODS AND RESULTS: Data were analyzed for 900 randomized patients with an ejection fraction

Surgical management of the patient with an implanted cardiac device: implications of electromagnetic interference.

OBJECTIVE: To identify the sources of electromagnetic interference (EMI) that may alter the performance of implanted cardiac devices and develop strategies to minimize their effects on patient hemodynamic status. SUMMARY BACKGROUND DATA: Since the development of the sensing demand pacemaker, EMI in the clinical setting has concerned physicians treating patients with such devices. Implanted cardiovertor defibrillators (ICDs) and ventricular assist devices (VADs) can also be affected by EMI. METHODS: All known sources of interference to pacemakers, ICDs, and VADs were evaluated and preventative strategies were devised. RESULTS: All devices should be thoroughly evaluated before and after surgery to make sure that its function has not been permanently damaged or changed. If electrocautery is to be used, pacemakers should be placed in a triggered or asynchronous mode; ICDs should have arrhythmia detection suspended before surgery. If defibrillation is to be used, the current flow between the paddles should be kept as far away from and perpendicular to the lead system as possible. Both pacemakers and ICDs should be properly shielded if magnetic resonance imaging, positron emission tomography, or radiation therapy is to be used. The effect of EMI on VADs depends on the model. Magnetic resonance imaging adversely affects all VADs except the Abiomed VAD, and therefore its use should be avoided in this population of patients. CONCLUSIONS: The patient with an implanted cardiac device can safely undergo surgery as long as certain precautions are taken.

Coronary perfusate composition influences diastolic properties, myocardial water content, and histologic characteristics of the rat left ventricle.

BACKGROUND: Recent studies found that edema, histology, and left ventricular diastolic compliance exhibit quantitative relationships in rats. Edema due to low osmolarity coronary perfusates increases myocardial water content and histologic edema score and decreases left ventricular filling. The present study examined effects of perfusate osmolarity and chemical composition on rat hearts. METHODS: Arrested American Cancer Institute (ACI) rat hearts (4 degrees C) were perfused with different cardioplegia solutions, including Plegisol (289 mOsm/L), dilute Plegisol (172 mOsm/L), Stanford solution (409 mOsm/L), and University of Wisconsin solution (315 mOsm/L). Controls had blood perfusion (310 mOsm/L). Postmortem left ventricular pressure-volume curves and myocardial water content were measured. After glutaraldehyde or formalin fixation, dehydration, and paraffin embedding, edema was graded subjectively. RESULTS: Myocardial water content reflected perfusate osmolarity, being lowest in Stanford and University of Wisconsin solutions (p<0.05 versus other groups) and highest in dilute Plegisol (p<0.05). Left ventricular filling volumes were smallest in dilute Plegisol and Plegisol (p<0.05). Osmolarity was not a major determinant of myocardial edema grade, which was highest with University of Wisconsin solution and dilute Plegisol (p<0.05 versus other groups). CONCLUSIONS: Perfusate osmolarity determined myocardial water content and left ventricular filling volume. However, perfusate chemical composition influenced the histologic appearance of edema. Pathologic grading of edema can be influenced by factors other than osmolarity alone.

Attenuation of reperfusion injury with probucol in the heterotopic rat cardiac isograft.

BACKGROUND: We tested the hypothesis that pretreatment with the antioxidant probucol attenuates reperfusion-induced diastolic abnormalities in the heterotopic rat cardiac isograft. METHODS: American Cancer Institute rats (n = 48) were divided into 6 groups. Hearts were arrested by coronary perfusion with 3 ml 4 degrees C University of Wisconsin solution at 60 mmHg. Eighteen donor hearts were divided into 3 groups of 6 and arrested either 1 hour after intraperitoneal injection of 3 ml oil with (Prob Tx) or without (Oil Tx) probucol (300 mg/kg) or without injection (Ctrl Tx). After a 90 minute storage period, abdominal isografting was performed with a total ischemic time of 2 hours. Following 15 minutes of blood reperfusion, donor hearts were rearrested and excised. Recipients' native hearts (NH, n = 18) were also arrested. Two additional groups with (Prob NR, n = 6) and without (Ctrl NR, n = 6) probucol pretreatment were arrested and subjected to 2 hours of ischemia without reperfusion. Postmortem LV pressure-volume curves and myocardial water content (MWC) were measured. RESULTS: At each pressure interval normalized LV volume (LVV) was significantly greater for Prob Tx than Oil Tx or Ctrl Tx. All isograft groups had significantly lower LVV at all pressure intervals and higher MWC than non-transplanted hearts. CONCLUSIONS: Pretreatment with probucol attenuates reperfusion-induced decreases in LVV in the heterotopic rat heart isograft model. Probucol, which is orally active in humans, merits further study for its potential to improve myocardial protection during cardiac surgery.

Pressure volume curves in arrested heterotopic rat heart isografts: role of improved myocardial protection.

BACKGROUND: To minimize decreases in left ventricular (LV) compliance immediately after rat heart transplantation, we tested several different methods of myocardial protection. MATERIALS AND METHODS: Five groups of ACI rat hearts (n = 6 each) were arrested by coronary perfusion with 5 ml of UW (University of Wisconsin), UW-BDM (UW with 2,3-butanedione monoxime), CU (Columbia University), or CU-BDM solution or by LV injection of potassium chloride and Ringer's lactate immersion (KCl/RL). After abdominal isografting and blood reperfusion for 15 min, transplanted hearts (TxH) were arrested and excised. Diastolic LV pressure-volume curves (LVPVCs) were correlated with myocardial water content (MWC). Native hearts (NH) were arrested identically to TxH and maintained at 4 degrees C by immersion. LVPVCs were measured at 15-min intervals for 90 min. RESULTS: In three of four pressure intervals at Time 0, normalized LV volume (LVV) was smaller (P < 0.05, ANOVA) in KCl/RL native hearts than in the four perfusion groups. LVV decreased significantly in NH after 45-75 min; LVV decreased similarly with time in all groups. In TxH, postarrest LVVs were higher with UW-BDM, CU-BDM, and CU than with UW or KCl/RL (P < 0.05, ANOVA). Expressing LVV of TxH as a percentage of NH, UW-BDM, CU, and CU-BDM provided qualitatively better diastolic properties than KCl/RL and UW. CONCLUSIONS: Thus rat LVPVCs can be improved after heart transplantation with alternative strategies of myocardial protection. KCl arrest decreases LV filling volume in this model and should be avoided.

Unconventional approaches to cardiac pacing in patients with inaccessible cardiac chambers.

BACKGROUND: Transvenous endocardial implantation can be impossible or contraindicated in patients with inaccessible right cardiac chambers. These patients usually undergo epicardial implantation, which has been associated with frequent rising thresholds and limited lead survival. We have used the following two alternative approaches in these patients: (1) transatrial puncture and passage of pacing leads for patients with no access to the right atrium and (2) ventricular pacing from the coronary sinus or its tributaries for patients with inaccessible ventricles. METHODS. We retrospectively reviewed our experience in 9 patients who had those procedures. Five patients had pacing from the coronary sinus, and 4 by transatrial puncture. RESULTS: Seven of the 9 patients had DDD pacing. Low acute pacing thresholds and satisfactory sensing levels were obtained with both approaches. One instance of high stimulation threshold (20%) occurred in the coronary sinus group and none in the transatrial puncture group. One patient in the transatrial puncture group died from unrelated causes. No malignant arrhythmias, pneumothorax, diaphragmatic pacing, or infectious complications have been observed. CONCLUSION: These unconventional approaches are safe, relatively simple, and reliable. Although the short-term follow-up is favorable, long-term follow-up is necessary to ascertain the relative merit of these approaches.

Conductance artifacts in a novel in vitro model of ventriculothoracic electrical coupling.

The utility of open chest conductance (COND) ventriculography is limited by artifacts altering the relationship between COND and left ventricular (LV) volume. Pressure-COND loops often lean to the left during LV volume reduction by caval occlusion. Time varying alterations in the pericardial-LV contact area affect electrical coupling in the open chest during the cardiac cycle, producing COND artifacts. In this study, an open-mediastinum model was constructed. Components represented the LV, blood, pericardium, and thoracic contents. Varying ventriculothoracic coupling was simulated by changing the volume of pericardial saline (0, 30, 60 ml). Raw dual field COND was repeatedly (n = 20) compared with volumes of normal saline from 60 to 120 ml at 5 ml intervals. Groups were compared by linear regression and repeated measures ANOVA. Artifacts significantly (p < 0.01) altered parallel COND, indicated by the y-intercept, with the exception of 0 versus 30 ml. The slope constant also changed significantly, with the exception of 30 versus 60 ml. These results suggest that variable pericardial-LV contact can cause time varying artifacts in COND in the open chest. Therefore, posterior insulation may reduce artifacts in COND ventriculography and should be tested for this effect.

Effect of implantable cardioverter-defibrillator implantation on surgical morbidity in the CABG Patch Trial. Surgical Investigators of the Coronary Artery Bypass Graft Patch Trial.

BACKGROUND: The Coronary Artery Bypass Graft (CABG) Patch Trial tested the hypothesis that prophylactic insertion of an implantable cardioverter-defibrillator (ICD) improves survival rates after high-risk CABG. We compared group-specific perioperative morbidity and mortality rates. METHODS AND RESULTS: Patients were randomized intraoperatively to undergo CABG (control subjects, n = 454) or CABG plus ICD implantation (n = 446). There were no significant differences between groups in the incidence of diabetes, ejection fraction < 0.25, end-diastolic pressure, prior myocardial infarction, or congestive heart failure. Cardiopulmonary bypass time averaged 106 minutes in control subjects and 127 minutes in the ICD group. At the inception of the trial, investigators were concerned that ICD therapy could increase surgical mortality rates or the incidence of shock, bleeding, congestive heart failure, arrhythmias, or deep sternal wound infection. Of these, only sternal wound infection was significantly more frequent in the ICD group (2.2% versus 0.4%, P < 0.05). Also more common in the ICD group were infection at a wound or catheter site (12% versus 6%), urinary tract infection (4% versus 1%), pneumonitis (8% versus 4%), respiratory insufficiency (13% versus 8%), transient central nervous system deficit (6% versus 2%), and psychotic reaction (4% versus 1%). The all-cause death rate was 6.7% in the ICD group and 4.6% for control patients (P = NS) at the time of the last surgical death, postoperative day 48. CONCLUSIONS: Epicardial ICD insertion during CABG is associated with an increase in perioperative infection. Although reporting bias may have influenced the data, if ICD insertion is indicated in CABG patients, metachronous endocardial implantation should be considered.

Intraoperative echocardiography: interpretation of changes in left ventricular wall thickness.

Quantitative two-dimensional echocardiography (Q2-DE) may be used to detect intraoperative changes in left ventricular (LV) mass (M) and wall thickness (h). Potential causes of change in h include physiological redistribution of myocardium, myocardial edema, reactive hyperemia, and intramyocardial hemorrhage. Changes in h, in the absence of changes in LV shape and volume, generally indicate increased LVM. When changes in h are accompanied by changes in shape or volume, changes in LVM can only be detected by mathematical modeling, unless the direction of the observed changes is opposite that expected with physiological redistribution. Histological observations essential to understanding current mathematical models are presented and related to the inherent solid geometry. Technical considerations in determination of LV mass by Q2-DE are discussed. New procedures that alter LV volume and geometry, such as the Batista operation, defy modeling by conventional methods. Modeling techniques that allow an experimental approach to understanding LVM and h under such conditions are presented.

Outpatient pacemaker procedures in orally anticoagulated patients.

Reexamination of surgical practices in the present era of cost containment has led to increased outpatient procedures including pacemaker surgery. While the safety and economic benefits of outpatient pacemaker surgery in nonanticoagulated patients is well documented, results of pacemaker operations in patients maintained on coumadin for thromboembolic prophylaxis have not been evaluated. In patients where complications with pacemaker surgery appeared successive, we have established a low incidence of complications. Recently, we extended this approach to the outpatient setting; this report retrospectively reviews our 4-year experience. During the study period, 150 patients underwent outpatient pacemaker procedures, including 37 patients receiving oral warfarin. There was no difference in the incidence of wound related and wound unrelated complications between patients receiving warfarin and the nonanticoagulated cohort. In addition, no wound hematomas, blood transfusions, or clinically significant bleeding episodes were noted among warfarin recipients. We conclude that pacemaker surgery in patients receiving oral anticoagulation is safe and feasible. The use of the cephalic cutdown technique avoiding blind subclavian punctures, meticulous attention to pocket hemostasis, and the use of small caliber unipolar positive fixation leads appears warranted in this selected group of patients at high risk for perioperative bleeding.

Effect of improved myocardial protection on edema and diastolic properties of the rat left ventricle during acute allograft rejection.

BACKGROUND: Studies of myocardial edema and diastolic dysfunction in rat heart transplantation have been flawed by ischemic injury. This study uses improved methods to prevent ischemic contracture. METHODS: Hearts of 30 ACI rats were transplanted into the abdomen of Lewis rats by use of cold University of Wisconsin solution for improved preservation. Left ventricular diastolic properties were expressed as volume at standardized pressure intervals. RESULTS: On posttransplantation day 3, mean left ventricular volume at 15 mm Hg in allografts (290 +/- 9 microl, SEM) was not significantly different vs isografts (299 +/- 32 microl), allografts on day 0 (337 +/- 28 ml) or day 1 (324 +/- 20 microl), or native hearts (334 +/- 19 microl). However, volume was reduced to 173 +/- 17 microl on day 4 and to 70 +/- 23 microl on day 5 (p < 0.05). Similar findings were obtained for volume at 5 and 10 mm Hg. Allograft myocardial water content on day 3, 76.3% +/- 5%, similar to allografts on day 0 and 1 and to isografts on day 3, increased to 77.6% +/- 8% on day 4 (NS) and 79.4% +/- 6% on day 5 (p < 0.05 vs day 0). Histologically, rejection in allografts was mild on day 3, moderate on day 4, and severe on day 5. CONCLUSIONS: Reduced left ventricular filling volume during rejection is only partially explained by edema. Abnormalities of diastolic properties previously attributed to the unloaded state of nonworking heart models may actually reflect inadequate peritransplantation myocardial protection.

Reversal of iatrogenic myocardial edema and related abnormalities of diastolic properties in the pig left ventricle.

OBJECTIVE: This study examines the resolution of iatrogenic edema and related changes in systolic and diastolic properties in the intact pig left ventricle. METHODS: The coronary arteries were perfused for 50 to 60 seconds with diluted blood (hematocrit value 10% +/- 1%, edema group, n = 5) or whole blood (hematocrit value 28% +/- 1%, control group, n = 6) infused into the aortic root during aortic crossclamping in conditioned, anesthetized pigs. After whole blood reperfusion, preload reduction by vena caval occlusion was used to define systolic and diastolic properties at 15-minute intervals. Left ventricular pressure and conductance, aortic flow, and two-dimensional echocardiography were recorded. RESULTS: Left ventricular mass (wall volume) in the edema group increased significantly compared with that in control pigs after crossclamp removal. Mass returned to preperfusion levels after 45 minutes. The ventricular stiffness constant (beta) increased significantly in the edema group versus the control group, returning to baseline by 30 minutes. The diastolic relaxation constant (tau) and base constant (alpha) did not differ between groups. There was no significant change in contractility. CONCLUSION: Increases in left ventricular mass and diastolic stiffness induced by coronary perfusion with hemodiluted blood resolve after 45 minutes of whole blood perfusion in pigs. This study defines physiologic effects of edema in the normal heart while eliminating most common confounding experimental errors.

A method for detecting changes in left ventricular mass during variations in filling volume.

Two-dimensional echocardiography has been useful for measuring changes in left ventricular mass (LVM) at constant left ventricular end-diastolic volume (LVEDV). Two-dimensional echocardiographic measurement of LVM changes during variations in LVEDV requires definition of the LVM/LVEDV relation because two-dimensional echocardiographic measurements could be affected by asymmetrical redistribution of LVM. Echocardiography data were recorded during caval occlusions in pigs (n = 6). Results confirm that A(M) (left ventricular [LV] short-axis cross-sectional [SACS] wall thickness area), was inversely related to AL (LV SACS lumen area), the average relation being A(M) = -0.33 AL + 20 (r = 0.82 +/- 0.05 [SE]). In addition, we developed a model that computes normal relation between LV SACS wall thickness area (AMc) and LV SACS lumen area (ALc) over a physiologic range of LVEDVs based on a single end-diastolic two-dimensional echocardiographic SACS image. Each computed relation corresponds uniquely to an LVM (LVMc). Theoretically, a difference between AMc/ALc relation before an intervention and the computed relation after the intervention would indicate a change in LVM. To test the utility of this model, edema was induced in a second group of pigs (n = 6) by coronary hemodilution. Two conditions were tested: pre-edema and edema. Serial AMc/ALc and LVMc were computed. Pre-edema and edema AMc were compared at matched LV SACS end-diastolic areas (ALc = 15 cm2). Results showed a significant increase in LVMc (two-tailed p value < 0.05), as observed by two-dimensional echocardiography. We conclude that the A(M) and AL are inversely related. This relation is useful for detecting alterations in LVM during variations in LVEDV.

Inhaled nitric oxide is not a myocardial depressant in a porcine model of heart failure.

BACKGROUND: Inhaled nitric oxide has been shown to be a potent and selective pulmonary vasodilator. Reports of increases in left ventricular end-diastolic pressure and episodes of pulmonary edema during the clinical use of inhaled nitric oxide in patients with preexisting left ventricular dysfunction have raised concerns that this agent may have myocardial depressant effects. We therefore undertook a study of the effects of inhaled nitric oxide on myocardial contractility in a porcine model of ventricular failure and pulmonary hypertension. METHODS: After inducing heart failure in 10 pigs by rapid ventricular pacing, hemodynamic measurements and pressure-volume diagrams (by the conductance method) were obtained in six animals at baseline and during administration of inhaled nitric oxide at concentrations of 20 and 40 ppm. Myocardial contractile state was assessed by the end-systolic pressure-volume relationship and preload-recruitable stroke work, whereas diastolic function was measured in terms of the end-diastolic pressure-volume relationship and the pressure decay time constant T. RESULTS: Baseline hemodynamics reflected heart failure and pulmonary hypertension, and inhaled nitric oxide induced significant reductions in mean pulmonary artery pressure and pulmonary vascular resistance. Although left ventricular end-diastolic pressure increased during administration of inhaled nitric oxide, no changes were observed in measures of systolic or diastolic function. CONCLUSIONS: Inhaled nitric oxide reduced pulmonary vascular resistance but did not alter myocardial contractility or diastolic function. Increases in left ventricular end-diastolic pressure during inhaled nitric oxide therapy are therefore not due to myocardial depression and may be related to increases in volume delivery to the left side of the heart resulting from reduced pulmonary vascular resistance.

Retrograde coronary perfusion: effects on iatrogenic edema and diastolic properties.

BACKGROUND: The relative merits of antegrade infusion and retrograde infusion of cardioplegic solution in terms of heart weight, myocardial water content, and ventricular diastolic properties are undefined. Accordingly, we compared antegrade and retrograde flow of hemodiluted blood in isolated, hypothermic porcine hearts. METHODS: After cardiectomy, 1 L of cold heparinized blood diluted with lactated Ringer's solution to concentrations ranging from 100% lactated Ringer's to 50% lactated Ringer's and 50% blood was perfused in an antegrade (n = 6) or retrograde (n = 6) fashion at mean pressures of 62 +/- 2 mm Hg (+/- standard error of the mean) and 49 +/- 2 mm Hg, respectively. Heart weight, myocardial water content, and left ventricular pressure-volume relationships were obtained before and after perfusion. RESULTS: In the comparison of measurements before and after perfusion, changes in heart weight (36 +/- 4 g versus 5 +/- 2 g; p < 0.05), myocardial water content (6.9% +/- 1.0% versus 2.5% +/- 0.4%; p < 0.01), and ventricular filling measured by normalized left ventricular volume at 10, 15, and 20 mm Hg were greater in the antegrade group. CONCLUSIONS: In the isolated porcine heart, retrograde flow is distinguished from antegrade flow by less change in heart weight and myocardial water content and no diastolic dysfunction.

Diastolic properties, myocardial water content, and histologic condition of the rat left ventricle: effect of varied osmolarity of a coronary perfusate.

BACKGROUND: Although myocardial edema is known to impair diastolic filling of the left ventricle, the interrelation of edema, histologic condition, and function has not been quantitated sufficiently for extrapolation to studies of multifactorial influences on diastolic properties. METHODS: Accordingly, ACI rat hearts arrested at 4 degrees C underwent coronary artery perfusion with a cardioplegia solution that was either unaltered (288 mOsm/L, P288 group, n = 6), diluted (144 mOsm/L, P144 group, n = 6), or concentrated (380 mOsm/L, P380 group, n = 6). Postmortem left ventricular pressure-volume curves and myocardial water content were measured. Myocardial samples were fixed in varying dilutions of glutaraldehyde. After dehydration and paraffin embedding, edema was graded subjectively (0 to 5), and myocardial interstitial spaces were determined by use of a semiquantitative method. RESULTS: Mean normalized left ventricular filling volume at 20 mm Hg filling pressure in the P144 group, 189 +/- 16 microliters (SEM), was reduced versus both the P288 (278 +/- 26 microliters) and the P380 (332 +/- 18 microliters) groups (p < 0.05, ANOVA). Mean myocardial water content in the P144 group, 80.7% +/- 1%, was increased versus the P380 (76.7% +/- 0.4%, p < 0.05) but not versus the P288 group (78.4% +/- 0.8%). In hearts preserved with 2.5% glutaraldehyde, mean edema grade and interstitial space in the P144 group (4.0 +/- 0.3) were increased versus the P380 (1.8 +/- 0.3, p < 0.05) but not the P288 group (2.7 +/- 0.5). Derived linear regressions relate water content to filling volume and histologic condition. CONCLUSIONS: Coronary perfusate osmolarity is thus associated with predictable changes in myocardial water content, left ventricular filling volume, and edema. These correlations allow definition of new hypotheses for the study of cardiac allograft rejection in patients and experimental animals.