Normothermic Continuous Antegrade Blood Cardioplegia Does Not Prevent Myocardial Edema and Cardiac Dysfunction
Background Normothermic continuous blood cardioplegia (BC) has been proposed to completely protect the myocardium during cardiac surgery. However, previous work from our laboratory suggests that BC could cause myocardial edema that produces cardiac dysfunction. The purpose of this present study was to evaluate the impact of BC on myocardial fluid balance and left ventricular function.
Methods and Results In 11 dogs, myocardial water content (MWC) was determined by microgravimetry. Myocardial lymph flow rate was measured after cannulation of the major prenodal cardiac lymphatic. Preload recruitable stroke work (PRSW) was calculated by sonomicrometry and micromanometry. The dogs were placed on normothermic cardiopulmonary bypass (CPB), and BC was delivered at either 80 to 90 mm Hg (BChigh; n=6) or 40 to 50 mm Hg (BClow; n=5) for 1 hour. Coronary sinus lactate and oxygen saturation monitoring demonstrated ischemia avoidance. BC was associated with substantial myocardial lymph flow rate decrease (P<.05) and myocardial edema development in both groups. MWC increased from 76.0±1.9% to 79.2±1.7% (P<.05) after 10 minutes of BChigh and from 75.9±0.6% to 78.9±1.4% (P<.05) after 30 minutes of BClow. PRSW decreased to 63±19% (BChigh) and 69±15% of control (BClow) at 30 minutes after CPB (P<.05). Myocardial lymph flow rate increases of threefold to fourfold that of control (P<.05) resulted in significant myocardial edema reduction associated with PRSW improvement to 71±17% (BChigh) and to 78±11% (BClow) at 2 hours after CPB.
Conclusions We conclude that BC is associated with compromised cardiac function despite ischemia avoidance. This cardiac dysfunction is due to myocardial edema caused by the combination of increased myocardial microvascular fluid filtration and decreased myocardial lymph flow rate during BC.
Normothermic continuous blood cardioplegia has resulted in substantial interest in its use as an alternative strategy for myocardial protection during cardiac surgery.1 2 3 4 In contrast to conventional hypothermic ischemic cardioplegia techniques, BC continuously provides oxygenated blood to the arrested myocardium, thus maintaining aerobic myocyte metabolism.2 3 Consequently, BC should completely preserve myocardial function because ischemia-induced myocardial stunning and necrosis are avoided. However, several studies have demonstrated no advantage of BC compared with intermittent hypothermic crystalloid or blood cardioplegia techniques.5 6 7 In a study of acute regional myocardial ischemia, BC prolonged contractile function recovery compared with intermittent hypothermic blood cardioplegia.8 These data suggest that BC may not completely preserve myocardial function.
In our previous study,9 conventional cold crystalloid cardioplegic arrest resulted in significantly decreased LV contractility associated with myocardial edema formation. We showed that impaired interstitial fluid removal contributed to this myocardial edema and cardiac dysfunction because we found myocardial lymph flow cessation in the arrested heart.9 We speculated that impaired lymph flow also would occur during cardiac arrest induced by BC, thus enhancing edema formation and LV dysfunction.
Myocardial edema during BC may develop for the following reasons. First, the arrested heart remains in diastole, which prolongs the time available for myocardial microvascular fluid filtration.10 Second, myocardial lymph drainage may be impaired because of the absence of rhythmic cardiac contraction. We previously demonstrated that organized ventricular contraction is the major determinant of myocardial lymphatic function.9 Thus, the combination of increased myocardial microvascular fluid filtration and decreased cardiac lymph drainage may cause myocardial edema and myocardial dysfunction induced by BC.
We hypothesized that BC causes myocardial dysfunction due to myocardial edema. The purpose of the present study was to evaluate the impact of BC on myocardial fluid balance, MWC, and LV function.
Myocardial tissue water content is an important variable in myocardial fluid balance evaluation. Conventional gravimetric methods require too large a sample for sequential determinations. In this study, we modified a microgravimetric technique for MWC quantification that was originally developed for measurement of cerebral edema.11 12 This technique allows determination of changes in MWC over time. Thus, the microgravimetric technique allows time course measurement of both myocardial edema development and resolution.
All procedures were approved by the University of Texas Animal Welfare Committee and were consistent with the National Institutes of Health’s “Guide for the Care and Use of Laboratory Animals.” Eleven conditioned mongrel dogs of either sex (mean±SD weight, 26.9±1.8 kg) were anesthetized with intravenous administration of 25 mg/kg thiopental sodium, intubated, and mechanically ventilated with 100% oxygen using a volume-cycled respirator (Siemens-Elema AB). We maintained anesthesia with intravenous infusion of 1% thiopental sodium in Ringer’s solution.
Subcutaneous needle electrodes were used to monitor the heart rate. We placed fluid-filled catheters into the left femoral artery and vein for mean arterial pressure monitoring and arterial blood sampling and fluid administration, respectively. We inserted a 7F Swan-Ganz thermodilution catheter into the pulmonary artery via the left jugular vein for central venous pressure, PAP, and CO determinations. In nine of the dogs we also advanced a 5F catheter via the right jugular vein into the coronary sinus for coronary sinus blood sampling. We then exposed the right femoral artery for subsequent CPB cannulation. After a median sternotomy, we incised the pericardium and placed a 1/8-in umbilical tape around the inferior vena cava to manipulate cardiac preload. Sonomicrometry crystals (5 MHz, Triton Technology Inc) were placed in the LV subendocardium across the septum–free-wall axis of the LV. A micromanometer-tipped pressure transducer (Millar Instruments Inc) was introduced into the LV cavity through the apex.
In eight dogs, we injected 0.2 to 0.5 mL of Evans blue dye solution (T-1824) into the LV myocardium to facilitate identification of the cardiac lymphatics. We cannulated the prenodal major left cardiac lymph trunk using a heparinized 20-gauge cannula (Medicut, Sherwood Medical Industries) as previously described.10 13 Myocardial lymph flow rate was measured with a calibrated pipette held at heart level. The resistance of this lymph cannula system was 4.11×10−3 mm Hg · min · μL−1.
Hemodynamic Parameters and LV Function
We connected the pressure monitoring catheters to pressure transducers (Isotec, Healthdyne Cardiovascular Inc), and data were recorded on an eight-channel strip-chart recorder (Grass Instrument Co). We determined CO in duplicate by injecting 10 mL ice-cold Ringer’s solution. LV pressure was measured with the micromanometer, and LV septum–free-wall diameter (dLV, in millimeters) was obtained with a sonomicrometer (Triton Technology Inc). These data were recorded at a frequency of 200 Hz (MacLab, World Precision Instruments Inc) and stored on a personal computer (Macintosh Quadra 700, Apple Computer Inc). If we assume a spherical shape, the LV volume (VLV [in milliliters]) can be calculated from the following equation9 :
where rLV is the radius of the LV in centimeters. Replacing rLV with (10−1 · dLV/2) (in millimeters) yields
We recorded LV pressure-volume loops during a 30-second period of inferior vena caval occlusion, and the PRSW (in mm Hg), an index of LV contractility, was calculated as the slope of the relation between LV end-diastolic volume and LV stroke work (SW).14 SW (in mL · mm Hg) was calculated as
where SV is the LV stroke volume, MEP is the LV mean ejection pressure, and EDP is the LV end-diastolic pressure. Ejection onset was defined at 10 ms after the time of +dP/dtmax, and end ejection was defined at the time of −dP/dtmax.14 The unstressed end-diastolic LV volume (V0, in milliliters) was defined as the x intercept of the PRSW relation.14
From the LV pressure signal we derived τ, an index of LV isovolumic relaxation, using the procedure described by Weisfeldt et al.15 Beginning at −dP/dtmax (end ejection), plotting dP/dt versus LV pressure yields a line with a slope of −1/τ whose negative reciprocal is τ (in milliseconds).15 This method of τ determination does not depend on the assumption that the LV pressure asymptote equals zero.15 16
For MWC determination, we modified a gravimetric technique that was originally developed for measurement of cerebral edema.11 12 MWC is determined by specific density measurement of small myocardial samples using a linear density gradient. Knowing the specific density of a myocardial sample, the percent gram of water per gram of tissue can be calculated.11 12 For preparation of the density gradient, we used two mixtures of kerosene (specific gravity, 0.773) and bromobenzene (specific gravity, 1.484). The specific gravities of these mixtures were adjusted to 0.990 and 1.080, respectively,11 12 and the density column was generated by use of a gradient former (model GC-0971, Bethesda Research Laboratories). We then calibrated the gradient with various K2SO4 solutions having known specific gravities of 1.079, 1.072, 1.067, 1.044, 1.035, 1.031, and 1.027. We carefully placed 10-μL drops of the K2SO4 solutions in the gradient and recorded the equilibration depth after 1 minute. We then plotted equilibration depth versus specific gravity and confirmed the linearity of the gradient by linear least-squares regression analysis. The mean correlation coefficient (±SD) was .991±.003; n=11.
To determine the specific gravity of myocardium, we introduced a biopsy forceps (Cordis Corporation) transapically into the LV and collected myocardial samples (6 to 8 mm3). These samples were gently placed into the density gradient, and the equilibration depth was recorded after 1 min. The gram H2O per gram myocardium, or MWC, (%), can be calculated from the following equation11 :
where SGmyo and SGdry are the specific gravities of the myocardial tissue sample and of dry myocardium, respectively. At the end of the experiment, a final myocardial tissue density measurement was performed. We then euthanatized the dog with intravenous pentothal overdose and saturated potassium chloride and rapidly excised the heart. Both ventricles were then weighed, after which they were stored in an oven and dried to a constant weight at 60°C. We calculated SGdry from the following equation11 :
where W and D are wet and dry weights of both ventricles, respectively. We assumed that SGdry did not change over the experimental period. All MWC measurements were performed at least in duplicate.
To test for myocardial vascular volume changes associated with cardiac paralysis, we additionally measured MWC after euthanasia with potassium in five dogs. We found no difference in MWC before versus after cardiac arrest (−0.1±0.2%; P=.92).
CPB and Continuous Warm Blood Cardioplegia
After preparation, heparin (300 IU/kg) was given intravenously for systemic anticoagulation. Additional doses of 75 IU/kg heparin were administered every 60 minutes throughout the experiment. We introduced a 16F arterial perfusion cannula into the prepared right femoral artery. A two-stage (34F and 38F) venous cannula (model TAC2, DLP Inc) was placed into the right atrium and inferior vena cava. The LV was vented with a 12F catheter inserted via the left atrium. CPB was performed with three roller pumps for extracorporeal circulation, LV drainage, and suction, respectively. We primed the extracorporeal circuit and the membrane oxygenator (Capiox 320, Terumo Corporation) with 800 mL of Ringer’s lactated solution and 1000 IU of heparin. A rectal temperature probe was placed, and the body temperature was maintained at 37°C during extracorporeal circulation with a heat exchanger. We maintained CPB flow between 70 and 90 mL/kg per minute and systemic perfusion pressure between 50 and 80 mm Hg.
We prepared warm (37°C) BC using a commercially available system (Sorin Biomedical Inc) that was connected to a heat exchanger. Four parts oxygenated CPB circuit blood was mixed with one part crystalloid cardioplegia (Plegisol, Abbott Labs) containing either 130 or 26 mmol/L K+.1 2 Both crystalloid cardioplegias were connected to a Y tube leading to the mixing system. This facilitated switching from high to low K+ concentration. Since arterial K+ concentration was between 3.5 and 5.5 mmol/L, the resulting K+ concentrations in the 4:1 BC mixtures were ≈29 to 30 mmol/L in the high-K+ BC and ≈8 to 10 mmol/L in the low-K+ BC, respectively.2 We then placed an aortic root cannula with pressure monitoring line (model 23009, DLP Inc) into the ascending aorta. After aortic cross-clamping, we initially delivered high-K+ BC into the aortic root. As soon as cardiac arrest was achieved, we switched to low-K+ BC and continuously infused low-K+ BC throughout the 60-minute cardiac arrest period unless myocardial electrical activity necessitated temporary return to high-K+ BC. We delivered BC at an aortic root pressure of 80 to 90 mm Hg in six dogs (high BC pressure group) and at 40 to 50 mm Hg in five dogs (low BC pressure group). To avoid systemic hyperkalemia during the 60-minute period of BC infusion, we administered fuorsemide and/or glucose-insulin infusion if arterial K+ concentration exceeded 5.5 mmol/L.
After instrumentation, we recorded baseline measurements of CO, mean arterial pressure, PAP, central venous pressure, LV pressure-volume loops, and myocardial lymph flow rate. Two myocardial samples for MWC determination were collected as described above. Arterial, BC, and coronary sinus plasma samples were frozen at −20°C for later lactate determination by use of an enzymatic test (Sigma Diagnostics). We placed the dog on CPB and initiated cardiac arrest by BC perfusion as described above. We measured all variables at 10, 30, and 50 minutes during BC administration (10′BC, 30′BC, and 50′BC, respectively). After 60 minutes of cardiac arrest we stopped BC, removed the aortic cross-clamp, and weaned the dog from CPB. At 30 and 120 minutes after separation from CPB (30′p.CPB and 120′p.CPB, respectively) we repeated all measurements.
To determine the effect of crystalloid CPB priming-induced hemodilution on MWC, we sampled myocardial biopsies at 15 to 20 minutes after CPB initiation and before aortic cross-clamping in five dogs.
All data presented in the text and tables are mean±SD. Data presented in figures are mean±SEM. We examined the time courses of each measured parameter using ANOVA for repeated measures and the F test. Post hoc comparisons were performed with Student’s t test, with a Bonferroni correction for multiple comparisons. A value of P<.05 was considered significant.
Data from 11 dogs are presented. The mean duration of CPB was 111±16 minutes at a flow rate of 81±8 mL · kg−1 · min−1 in the high BC pressure group (n=6) and 115±9 minutes at 77±6 mL · kg−1 · min−1 in the low BC pressure group (n=5). Hemodilution due to the crystalloid CPB priming resulted in an MWC increase of 0.9% (from 75.9±0.6% at baseline to 76.8±0.9%; n=5; P=.055). During the 60-minute period of aortic cross-clamping, we perfused the hearts with 6208±2006 mL BC (418±169 mL high-K+ BC and 5790±1885 mL low-K+ BC) at an aortic root pressure of 86.7±2.6 mm Hg in the high BC pressure group. Hearts in the low BC pressure group were perfused with 2135±342 mL BC (464±263 mL high-K+ BC and 1671±171 mL low-K+ BC) at an aortic root pressure of 51.6±4.2 mm Hg. Tables 1⇓ and 2⇓ provide data on BC flow, coronary sinus oxygen saturation, and hematocrit, as well as myocardial lactate uptake rates (negative lactate uptake indicates lactate production) for both groups. We believe myocardial ischemia was avoided during BC because (1) arrested hearts did not exhibit significant myocardial lactate production and (2) coronary sinus oxygen saturation was always substantially higher than our baseline value of 40.3±6.7% (n=9; P<.001).
Sinus rhythm resumed spontaneously in all 11 dogs after cross-clamp removal, and no dog required positive inotropic support for weaning from CPB. Tables 3⇓ and 4⇓ show all hemodynamic variables, LV function data, and arterial hematocrit concentrations for both groups.
Compared with baseline, LV contractility as measured by PRSW was significantly decreased at 30 minutes after CPB in both groups. At 2 h after CPB, PRSW was still significantly lower compared with baseline; however, there was a trend for improved contractility compared with 30 minutes after CPB in both groups (P=.27 and P=.24, respectively). The time constant of isovolumic relaxation, τ, remained unchanged after CPB. Compared with the low BC pressure group, arterial hematocrit was significantly lower at 30′p.CPB and at 120′p.CPB due to the larger amounts of crystalloid cardioplegia administered in the high BC pressure group. For all the other variables, there were no significant differences between 30′p.CPB and 120′p.CPB.
Fig 1⇓ shows the changes in MWC induced by BC. In both groups, significant edema formed after only 10 min of BC perfusion and remained on the same level until 30′p.CPB. However, 2 hours after separation from CPB, part of the myocardial edema was resolved. Although MWC at 120′p.CPB was still higher compared with baseline in both groups, there was a significant decrease from 30′p.CPB to 120′p.CPB in the high BC pressure group and from 50′BC to 120′p.CPB in the low BC pressure group. There was no significant difference between the time courses of MWC for both groups.
Fig 2⇓ demonstrates the impact of BC on myocardial lymph drainage. Compared with baseline, myocardial lymph flow rate decreased significantly to <30% during BC and increased to threefold to fourfold that of control after separation from CPB in both groups.
Our data show that despite ischemia avoidance, BC does not completely preserve myocardial function. We found BC to be associated with myocardial edema development and compromised LV function. We believe that the combination of increased microvascular fluid filtration and insufficient myocardial lymph drainage in the arrested perfused heart during BC caused interstitial myocardial edema development and, thus, LV dysfunction during the period after BC.
Fluid movement out of the coronary microvascular exchange vessels is most likely enhanced during BC for the following reason. Diastole is the phase during which perfusion and filtration occur.10 As the arrested heart remains in a continuous diastolic state, time for filtration is increased. Thus, the time-dependent filtration coefficient, which represents water permeability,17 is increased during diastolic arrest because the entire “cardiac cycle” is available for filtration in the absence of systole. Assuming that systole represents about one third of the cardiac cycle,18 the filtration coefficient should be at least 1.5 times higher in the arrested heart compared with the normal beating heart.
In the normal beating heart, increased microvascular filtration causes increased myocardial lymph flow rate, which is regarded as an important “anti-edema safety factor”.10 13 17 19 20 During BC, however, we observed significant reduction of myocardial lymph flow rate (Fig 2⇑). This is in agreement with our previous work in which we demonstrated in a hypothermic cardioplegia model that organized ventricular contraction is the major determinant of myocardial lymph propulsion.9 Thus, the impaired cardiac lymph drainage due to the lack of rhythmic cardiac contraction and relaxation in combination with the increased microvascular filtration in the arrested heart causes myocardial fluid accumulation during BC. This is supported by Weng et al,21 who found that repeated blood perfusion of arrested hearts resulted in progressive heart weight increases.
Myocardial Fluid Balance During BC
Higher BC perfusion pressure most likely results in increased myocardial capillary pressure, thus increasing microvascular filtration rate. We chose BC pressures of 40 to 50 and 80 to 90 mm Hg to cover the range of clinically and experimentally applied BC techniques.1 2 5 7 8 21 22 The myocardial edema accumulation rate was not significantly different between the two groups, although edema tended to accumulate more slowly in the hearts perfused at 40 to 50 mm Hg BC pressure (Fig 1⇑). Since myocardial lymph flow was similar in both groups, fluid filtration must have been lower in the low BC pressure group at least during the first 10 minutes of BC.
From our data, we can determine the proportion of excess myocardial water due to increased microvascular filtration compared with that caused by decreased lymph flow. On the basis of myocardial wet and dry weights measured at the end of the experiments, we calculated that the observed MWC increase during the first 10 minutes of BC represents about 18 and 10 mL of additional myocardial water in the high and low BC pressure groups, respectively. Assuming that our measured myocardial lymph flow rate approximates 85% of the total cardiac lymph flow rate,9 we estimate that decreased myocardial lymph flow could account for only ≈1 mL of the excess fluid that accumulated in both groups during the first 10 minutes of BC. Thus, the major cause of myocardial fluid accumulation in the arrested perfused heart must have been an increased microvascular filtration rate. The magnitude of the increase in filtration rate can be demonstrated in the following fashion. Under baseline conditions, microvascular filtration rate equals total myocardial lymph flow rate, or ≈0.11 mL · min−1, by the above assumption concerning the percentage of total myocardial lymph flow we collect with our lymphatic. Net filtration rate can be estimated to be ≈1.7 mL · min−1 in the high BC pressure group during the first 10 minutes of BC [18 mL excess water · (10 min)−1−0.1 mL · min−1 decreased lymph flow], which is about 15-fold higher than baseline. Even at 50 mm Hg BC perfusion pressure, net filtration rate was still 10-fold higher than baseline. This clearly demonstrates the physiological factor, increased filtration, that predisposes hearts receiving BC to develop edema.
Surprisingly, we found that myocardial edema did not worsen during the second 30 minutes of BC despite continuous coronary perfusion (Fig 1⇑). Since myocardial lymph flow rate remained low during the whole BC period (Fig 2⇑), the absence of further myocardial fluid accumulation after 30 minutes of BC could be due to decreased microvascular fluid filtration rate and/or increased interstitial fluid removal via alternative pathways, including epicardial transudation. Myocardial microvascular fluid filtration decreased after 30 minutes of BC, probably by the following mechanism. Myocardial edema formation during the first 30 minutes of BC was probably accompanied by a progressive myocardial interstitial pressure increase. Thus, the hydrostatic pressure gradient determining the rate at which fluid leaves the microvascular exchange vessels17 23 most likely decreased due to myocardial edema, thereby decreasing myocardial fluid filtration. This is supported by previous work that demonstrated that myocardial interstitial pressure increased because of myocardial edema10 13 and that increased myocardial interstitial pressure was associated with decreased myocardial microvascular fluid filtration.24
Myocardial Edema Resolution
Total myocardial lymph flow at 120′p.CPB is sufficient to remove the excess fluid in only 1 hour. However, actual edema resolution rates were only 20% and 30% per hour in the high and low BC pressure groups, respectively. The most likely reason for this discrepancy is increased microvascular fluid filtration owing to increased microvascular permeability. This is supported by several studies that demonstrated CPB-induced activation of numerous mediators that are capable of producing increased vascular permeability.25 26 27 Thus, myocardial lymph flow data alone overestimate edema resolution rate because the observed myocardial lymph flow increase after CPB probably reflects the combination of myocardial edema resolution and increased microvascular fluid filtration. Nevertheless, we found that increased myocardial lymph flow was associated with significant MWC reduction at 2 hours after separation from CPB. This demonstrates the important role of cardiac lymph drainage for myocardial edema resolution.
Impact of Myocardial Edema on LV Function
The impact of myocardial edema on LV function is demonstrated in Fig 3⇓. We found a direct inverse relation between MWC and LV contractility as measured by PRSW. Each percentage increase in MWC was associated with an 11% contractility decrease. As shown in Fig 3⇓, LV contractility was depressed by ≈35% at 30 minutes after CPB, but the improvement 90 minutes later was associated with significant myocardial edema resolution (Fig 1⇑). This is in agreement with other studies that demonstrate the direct impact of myocardial edema on LV function. Laine and Allen13 showed a 30% decrease in cardiac reserve in dogs with myocardial edema similar to that found in the present study. More specifically, Davis et al28 produced LV edema in an acute dog model and showed that edema leads to significant LV dysfunction similar to that found in the present study.
Although several investigators have demonstrated that myocardial edema is associated with cardiac dysfunction,9 13 28 29 30 the mechanisms by which myocardial edema compromises myocardial function are not completely understood. Edema could cause cardiac dysfunction by a biomechanical effect. Excess myocardial fluid has been shown to increase myocardial stiffness, thus decreasing LV compliance.21 31 The impaired LV compliance combined with the viscous effects of moving excess interstitial water could compromise the efficiency of myocardial contraction. This mechanism is supported by investigators who demonstrated increased myocardial energy requirements associated with edema.32 33 Furthermore, interstitial fluid accumulation expands the myocardial interstitium, increasing oxygen diffusion distances between capillaries and myocytes. This is particularly important in the heart because it operates at near-maximum oxygen extraction. Thus, interstitial myocardial fluid accumulation could cause decreased contractility by inducing myocyte ischemia.34
Our data show that myocardial edema resolution at 120′p.CPB (Fig 1⇑) was associated with a further increase in myocardial lymph flow rate compared with 30′p.CPB (Fig 2⇑). The only possible explanations for this increased lymph flow are a more increased filtration rate or improved cardiac lymphatic drainage. A more increased filtration rate seems to be unlikely because MWC decreased. We believe that the improved LV contractility at 120′p.CPB (Fig 3⇑) caused the observed myocardial lymph flow increase. This emphasizes the importance of regular myocardial contraction for sufficient cardiac lymphatic function, as demonstrated in our previous work.9
BC Versus Standard Hypothermic Cardioplegia
In our previous study of cold (4°C) crystalloid cardioplegic arrest, MWC increased to 78.5% at 1 hour after CPB,9 which is similar to the degree of edema we found with BC at 30 minutes after CPB. In contrast to BC, cold crystalloid cardioplegia resulted in both impaired isovolumic relaxation, as indicated by τ prolongation, and LV dilatation, as indicated by V0 increase.9 In Fig 4⇓, the low BC pressure group is compared with the conventional crystalloid cardioplegia group. In contrast to the BC group, which showed contractility recovery associated with edema resolution from 30 to 120 minutes after CPB, myocardial edema and depressed contractility persisted 1 hour after CPB in the crystalloid cardioplegia group. This suggests that myocardial protection using BC is superior to conventional crystalloid cardioplegia.
In conclusion, the results of our study demonstrate that fluid movement out of the myocardial capillaries into the interstitium is enhanced during BC. Continuous perfusion of the arrested heart resulted in myocardial edema development that was directly associated with compromised LV function. How could edema development in the arrested perfused heart be minimized? Addition of osmotic or oncotic active substances such as mannitol or albumin to the blood cardioplegia does not appear to reduce the microvascular filtration to a rate sufficient to prevent edema.21 30 This is most likely due to the myocardial microvascular exchange barrier’s large surface area,35 large pores,36 37 and high protein permeability.10 37 Reduction of BC perfusion pressure below 40 mm Hg may reduce microvascular filtration and limit edema formation but risk ischemia. The ideal BC perfusion pressure that minimizes microvascular fluid filtration and simultaneously ensures homogeneous myocardial perfusion to prevent ischemia has not been determined. Further improvement of myocardial protection regimens will require inclusion of myocardial fluid balance principles.
Selected Abbreviations and Acronyms
|BC||=||normothermic continuous blood cardioplegia|
|MWC||=||myocardial water content|
|30′p.CPB||=||30 minutes after cardiopulmonary bypass|
|120′p.CPB||=||120 minutes after cardiopulmonary bypass|
|PAP||=||pulmonary artery pressure|
|PRSW||=||preload recruitable stroke work|
|SGdry||=||specific gravity of the dry myocardium|
|SGmyo||=||specific gravity of the myocardial tissue sample|
|τ||=||time constant of left ventricular isovolumic relaxation|
This study was supported by National Heart, Lung, and Blood Institute grant HL-36115 and the American Heart Association. Dr Mehlhorn is the recipient of a fellowship granted by the German Research Foundation (Deutsche Forschungsgemeinschaft). The authors thank Mark Brown for his excellent technical assistance.
Guest editor for this article was David F. Torchiana, MD, Massachusetts General Hospital, Boston.
- Received February 23, 1995.
- Revision received March 29, 1995.
- Accepted April 16, 1995.
- Copyright © 1995 by American Heart Association
Yau TM, Weisel RD, Mickle DAG, Ivanov J, Mohabeer MK, Tumiati L, Carson S, Liechtenstein SV. Optimal delivery of blood cardioplegia. Circulation. 1991;84(suppl III):III-380-III-388.
Ko W, Zelano J, Isom OW, Krieger KH. The effects of warm versus cold blood cardioplegia on endothelial function, myocardial function, and energetics. Circulation. 1993;88(pt 2):II-359-II-365.
Lajos TZ, Espersen CC, Lajos PS, Fiedler RC, Bergsland J, Joyce LT. Comparison of cold versus warm cardioplegia: crystalloid antegrade or retrograde blood? Circulation. 1993;88(pt 2):II-344-II-349.
Mehlhorn U, Davis KL, Burke EJ, Adams D, Laine GA, Allen SJ. Impact of cardiopulmonary bypass and cardioplegic arrest on myocardial lymphatic function. Am J Physiol. 1995;268:H178-H183.
Laine GA, Granger HJ. Microvascular, interstitial, and lymphatic interactions in the normal heart. Am J Physiol. 1985;249:H834-H842.
Laine GA, Allen SJ. Left ventricular myocardial edema: lymph flow, interstitial fibrosis, and cardiac function. Circ Res. 1991;68:1713-1721.
Glower DD, Spratt JA, Snow ND, Kabas JS, Davis JW, Olsen CO, Tyson GS, Sabiston DC, Rankin JS. Linearity of the Frank-Starling relationship in the intact heart: the concept of preload recruitable stroke work. Circulation. 1985;71:994-1009.
Weisfeldt ML, Weiss JL, Frederiksen JT, Yin FCP. Quantification of incomplete left ventricular relaxation: relationship to the time constant for isovolumic pressure fall. Eur Heart J. 1980;1:119-129.
Braunwald E, Sonnenblick EH, Ross J Jr. Mechanisms of cardiac contraction and relaxation. In: Braunwald E, ed. Heart Disease. 4th ed. Philadelphia, Pa: WB Saunders Co; 1992:351-392.
Laine GA. Microvascular changes in the heart during chronic arterial hypertension. Circ Res. 1988;62:953-960.
Mehlhorn U, Davis KL, Laine GA, Allen SJ. Myocardial microvascular filtration is directly related to arterial blood pressure in anesthetized dogs. FASEB J. 1994;8:A1050. Abstract.
Weng ZC, Nicolosi AC, Detwiler PW, Hsu DT, Schierman SW, Goldstein AH, Spotnitz HM. Effects of crystalloid, blood, and University of Wisconsin perfusates on weight, water content, and left ventricular compliance in an edema-prone, isolated heart model. J Thorac Cardiovasc Surg. 1992;103:504-513.
Buckberg GD, Brazier JR, Nelson RL, Goldstein SM, McConnell DH, Cooper N. Studies of the effects of hypothermia on regional myocardial blood flow and metabolism during cardiopulmonary bypass, 1: the adequately perfused beating, fibrillating, and arrested heart. J Thorac Cardiovasc Surg. 1977;73:87-94.
Starling EH. On the absorption of fluids from the connective tissue spaces. J Physiol (Lond). 1896;19:312-326.
Laine GA, Williams JP, Allen SJ, Gabel JC, Drake RE. Regulation of transmicrovascular fluid and protein flux by interstitial fluid pressure. Fed Proc. 1987;46:1532. Abstract.
Dauber IM, Parsons PE, Welsh CH, Giclas PC, Whitman GJR, Wheeler GS, Horwitz LD, Weil JV. Peripheral bypass-induced pulmonary and coronary vascular injury. Circulation. 1993;88:726-735.
Davis KL, Mehlhorn U, Laine GA, Allen SJ. Myocardial edema, left ventricular function, and pulmonary hypertension. J Appl Physiol. 1995;78:132-137.
Laine GA, Allen SJ. Increased cardiac energy consumption accompanies myocardial interstitial edema. FASEB J. 1992;6:2038. Abstract.
Ziegler WH, Goresky CA. Transcapillary exchange in the working left ventricle of the dog. Circ Res. 1971;29:191-207.
Wearn JT. The extent of the capillary bed of the heart. J Exp Med. 1928;47:273-291.
Pilati CF. Macromolecular transport in canine coronary microvasculature. Am J Physiol. 1990;258:H748-H753.