Ann Thorac Surg 2001;71:1312-1319
© 2001 The Society of Thoracic Surgeons
Original article: cardiovascular
Laboratory confirmation of clinical heart allograft preservation variability
Todd L. Demmy, MDa,
Timothy A. Turpin, BSa,
Colette C. Wagner-Mann, PhDa
a Division of Cardiothoracic Surgery, University of Missouri-Columbia, School of Medicine, Columbia, Missouri, USA
Accepted for publication November 27, 2000.
Address reprint requests to Dr Demmy, Division of Cardiothoracic Surgery, University of Missouri-Columbia, MA 312 HSC, One Hospital Dr, Columbia, MO 65212
e-mail: demmyt{at}health.missouri.edu
 |
Abstract
|
|---|
Background. Previously, we reported survival differences from the national heart transplant registry favoring centers that used intracellular organ preservation solutions. To eliminate center selection bias, we tested some of these solutions in a biventricular working rat heart model to determine their relative efficacy.
Methods. Using 103 Sprague-Dawley rat hearts perfused with modified Krebs-Henseleit buffer, both ventricles functioned with adjustable independent preload and afterload and their pressure–length loops generated load-insensitive measurements of cardiac performance. After 15 minutes of stable function, each heart sustained 180 minutes of cold (4°C) ischemia after a 5-minute perfusion by University of Missouri (UMC), Plegisol, Collins, University of Wisconsin, Custodiol, or Roe solutions. Eighty-two hearts were reperfused and the remainder were used for ATP analyses.
Results. Although the extracellular solution Plegisol showed good recovery of traditional hemodynamic values, including developed pressure and cardiac output, intracellular solutions like Roe had superior preservation of load-insensitive indices such as preload recruitable stroke work: Roe (intracellular) 103% ± 13%; Custodiol (intracellular) 96% ± 9%; UW (intracellular) 69% ± 12%; Collins (intracellular) 68% ± 9%; Plegisol (extracellular) 68% ± 7%; and University of Missouri (extracellular) 56% ± 10% (p = 0.04). Furthermore, recovery with intracellular solutions tended to be gradual but more progressive after ischemia in contrast to an early plateau shown by extracellular (p < 0.001). Right ventricular recovery and ATP measurements were similar between groups.
Conclusions. These data support the superiority of certain intracellular preservation solutions and provide evidence that optimal heart organ protection may be difficult to judge clinically using hemodynamic values routinely available to the heart transplant surgeon. Care should be taken to verify the performance of some solutions used in heart organ transplantation.
|
Introduction
|
|---|
There is great variation in the practice of clinical heart organ preservation because of the availability of multiple intracellular (I)- and extracellular (E)-based solutions with documented laboratory and clinical efficacy, many undocumented custom variations of these solutions perceived successful in routine open heart operations at specific institutions, and a lack of consensus regarding this aspect of heart organ procurement [1, 2]. Suboptimal organ preservation is masked by routine ß-agonist use or other adverse donor or recipient characteristics. Furthermore, the need for donor heart tissue healing resulting from impaired organ preservation promotes rejection and may lead to manifestation of adverse effects beyond the initial hospitalization. We found impaired 1-year survival associated with the use of E organ preservation solutions by matching heart organ preservation solutions with 9,401 recipient patients [2]. It is unknown, however, whether to attribute these findings directly to the solution or to other factors that affect survival favorably at centers using I solutions.
Recently, we reported a biventricular working rat heart model that is able to provide complex hemodynamic data and can mimic the clinical heart transplant scenario [3, 4]. The purpose of this study was to compare the effectiveness of several organ preservation solutions in the biventricular working rat heart and relate them to our clinical findings.
|
Material and methods
|
|---|
Animals
Hearts from 103 Sprague-Dawley rats were used to compare the effects of ischemia for various organ preservation solutions. After ischemia, 82 were reperfused and the remainder were used for ATP measurements. All animals received humane care in compliance with the "Guide for the Care and Use of Laboratory Animals" published by the National Institutes for Health (NIH publication no. 85-23, revised 1985). After fasting, the animals were heparinized (sodium heparin, 2.5 mg) and anesthetized (chloral hydrate, 36 mg/100 g body weight) by intraperitoneal administration. The thorax was opened, and the heart was removed. The isolated heart was placed in a beaker of cold modified Krebs-Henseleit bicarbonate buffer (K-H; see Isolation and Perfusion of the Heart section) (2°C) before being used for the experimental study.
Perfusion apparatus for biventricular working hearts
The left working heart model was modified as diagrammed in our earlier study [3, 4]. We have since enhanced it to provide a compliance chamber for the right heart. This apparatus is comprised of six major parts: (1) cannula assembly, (2) compliance chambers (2), (3) venous-oxygenation reservoir, (4) heart chamber, (5) recirculation reservoir with filter, and (6) peristaltic pump. Compliance chambers were used for dynamic impedence instead of static overflow columns for right and left heart afterload. Reliable, rapid, linear changes in afterload can be brought about by adjusting pressure within these compliance chambers [3]. The cannulation assembly consists of a standard fixed left working heart apparatus that is modified to include flexible right atrial and pulmonary cannulae. These cannulae are tied to and rely on the contralateral inflow (left atrial) and outflow (aortic) cannulae for support.
Intraventricular pressure monitoring is achieved using Millar high-fidelity pressure transducers. After bilateral placement of apical ventricular cannulae with tapered extraventricular tips, 2F Millar pressure transducers are introduced into the cannulae and held snugly in place by the tapered tips (Fig 1).
View larger version (159K):
[in this window]
[in a new window]
|
Fig 1. Photograph of cannulated heart. (A = aortic cannula; C = left atrial cannula; D = pulmonary artery cannula; L = ventricular cannula; M = Millar cannula; N = crystals attached to ventricles.) See reference for schematic diagram and other information [3]. (Reprinted by permission of Journal of Biomedical Science/S. Karger AG.)
|
|
The cannulation assembly also supports 1-mm ultrasonic crystals (Crystal Biotech, Northborough, MA) used to measure changes in distance between crystal pairs placed on the epicardium. Crystal wire pairs are passed through 6-mm by 5-mm sections of soft tygon tubing that are secured to the cannulation assembly by malleable 20-gauge copper wire. The soft tubing prevents wire damage and dislodgement when compressed by a C-clamp.
The heights of venous oxygenation reservoirs were adjusted independently to produce the desired venous pressures or preload. All reservoirs were water jacketed, maintained at 37°C, and K-H buffer was equilibrated continuously with 95% O2 and 5% CO2. Preliminary retrograde perfusion and the recording of aortic pressure were accomplished through a side arm of the aortic cannula. Pulmonary artery and atrial pressures were measured with in-line transducers. The completed preparations and cannulae assemblies were fitted into water-jacketed heart chambers to maintain the selected temperature.
Isolation and perfusion of the heart
Under general anesthesia, a transverse upper abdominal incision was made. After bilateral ventrolateral thoracotomy, retraction and inversion of the ventral chest wall was achieved. The heart was isolated with sufficient vessels attached to it for subsequent cannulation and placed in a beaker of cold K-H (2°C) to stop its activity. After aortic cannulation, 15 minutes of preliminary retrograde perfusion at a perfusion pressure of 60 mm Hg was begun without recirculation. K-H at 37°C equilibrated with 95% O2 and 5% CO2 containing normal rat plasma levels of amino acids [5], 15 mmol/L glucose, and 400 µU/mL insulin was used. During preliminary perfusion, excessive tissue was trimmed, and the left atrium (through a pulmonary vein), right atrium (through the inferior vena cava), and pulmonary artery were cannulated in sequence.
Before insertion of the atrial cannulae, ventricular cannulae were inserted transatrially. The pulmonary artery cannula was flexible and freely movable to facilitate aortic and left atrial cannulation. A 4-0 silk ligature was passed around the distal portion of the pulmonary artery before the cannula was inserted. After cannulation of the heart, biventricular perfusion was started by clamping the retrograde perfusion inflow and opening the inflow tubing connected to the atrial cannulae.
Once stabilized, ultrasonic crystals were applied to the epicardium with cyanoacrylic glue. This was done by selecting the greatest short axis diameter and applying the crystals in ventral–dorsal orientation. A thin layer of glue was applied precisely to the crystal using a micropipette and an air jet removed excess moisture from the target area.
Measurements of cardiac hemodynamics and metabolism
Atrial, pulmonary arterial, aortic, and compliance chamber pressures were measured simultaneously using TRANSPAL pressure transducers (Abbott Laboratories, North Chicago, IL) and recorded with a multichannel data analyzing system (PO-NE-MAH) (Valley View, OH).
Aortic output and pulmonary flow (PF) were measured by electromagnetic flow probes on the outflow tracts. Coronary flow was measured from the pulmonary artery outflow and the right atrial inflow was temporarily interrupted.
Developed pressure was the difference between arterial systolic and diastolic pressures. Cardiac output (CO) was the sum of aortic output and coronary flow. The left ventricular stroke work index (LVSWI) and right ventricular stroke work index (RVSWI) were calculated by the following equations: LVSWI = (mean aortic pressure x CO x 14.4)(heart rate x body surface area); and RVSWI = (mean pulmonary artery pressure x PF x 14.4)(heart rate x body surface area).
Estimates of ventricular work were made using areas from the pressure–length loops. This was accomplished by integrating the pressure–length loop regions using PO-NE-MAH software. Also measured were percent systolic shortening and a power work integration [
(-dL/dt)(ventricular pressure)dt]. Estimates of ventricular elastance were made by capturing raw pressure–length loop data while decreasing right atrial and left atrial pressure. These data were transferred to a computerized database (Microsoft Visual Foxpro, Redmond, WA). Software developed in this laboratory was used to determine the area work of each loop (trapezoidal estimation) and select the end-systolic point(s) of each loop. End-systolic points for each loop were edited and confirmed manually. The resulting data points were fit linearly into a standard model of time varying elastance (Fig 2): Ees = P (t).
View larger version (49K):
[in this window]
[in a new window]
|
Fig 2. Elastance lines and loops in one heart. Broken line (loops not shown) shows the inversed slope (elastance) and x-intercept ("creep") of one heart subjected to ischemia. Corresponding right ventricular loops shown in insert. Linear regression analyses of points that comprise the base of these loops provide diastolic slope. This line’s slope (not shown) increases as ventricular compliance decreases. (LV = left ventricular.) (Reprinted by permission of Journal of Biomedical Science/S. Karger AG.)
|
|
Next, the area of each loop was calculated and plotted against the end-diastolic dimension for each loop. The slope of this line represented preload-recruitable stroke work that varied consistently and directly with ventricular performance (Fig 3).
View larger version (19K):
[in this window]
[in a new window]
|
Fig 3. Preload recruitable stroke work graph from one heart. In past ischemic experiments, the relation of loop area to end-diastolic dimension provided a more reproducible index of ventricular performance.
|
|
At a predetermined time under proper perfusion, each heart was frozen with Wollenberger tongs that had been cooled to the temperature of liquid nitrogen. The frozen heart pellets were pulverized before being extracted with 6% perchloric acid (0°C). After centrifugation, the supernatant was neutralized and filtered (0.2 µm) before ATP, ADT, AMP, creatine phosphate, and creatine content were determined using a high-performance liquid chromatography system (Waters Co, Milford, MA) with the method published previously [6]. The pellet was subjected to alkaline digestion before samples were taken for protein [7] measurements.
Unfortunately, a freezer malfunction prevented the batch analyses of ATP specimens from two groups (Roe, Custodiol) added subsequently to the original protocol.
Experimental design
One hundred three hearts were randomly assigned to be perfused with one of six different organ preservation solutions according to the day of the experiment and the investigator was blinded as to the identity of the solution. The solutions groups were originally University of Missouri (UMC, n = 21), Collins (n = 17), Plegisol (n = 14), and University of Wisconsin (UW, n = 16) [8–11]. The UMC solution is an extracellular cardioplegic solution used extensively both for clinical open heart operations as well as the early cardiac transplant experience at the University of Missouri. The constituents of this solution are: Na, 145 mEq/L; K, 24 mEq/L; Ca, 2.7 mEq/L; Cl, 109 mEq/L; lactate, 28 mEq/L; bicarbonate, 15 mEq/L, solumedrol, 250 mg/L; and dextrose, 12.5 g/L. The other solutions components are as referenced. Plegisol, Custodiol, and UW were obtained from the vendor, whereas Roe and Collins were prepared according to the cited references. The group sizes varied somewhat depending on the success of the preparation on the particular days. Two solutions were introduced into the protocol later. The first was Custodiol (n = 4, aka Bretschneider’s solution) because it was to become available as a proprietary alternative to Plegisol solution [12]. Initial results were so encouraging that it was advanced to 6-hour ischemia experiments currently in progress. Roe (n = 10) solution was the other late addition because of the favorable clinical effects associated with this solution found in the retrospective study [8].
After procurement, cannulation, and application of electronic crystals as described, the hearts were allowed to stabilize for 15 minutes while baseline measurements were taken. The left compliance chamber was set at 60 mm Hg afterload and the right chamber provided 10 mm Hg. Compliance chambers were lowered while loops were captured electronically. First right, then left loops were determined as routine.
Next, the aorta and pulmonary outflow were clamped and retrograde perfusion of 4°C organ preservation solution was initiated and continued for 5 minutes. The Millar catheters were immediately removed to allow venting of both ventricles and measurement of coronary perfusate flow. After perfusion, the heart remained nonperfused and immersed in 4°C K-H for 3 hours. Then the heart was reperfused with fresh oxygenated K-H in a nonworking retrograde manner. When ventricular fibrillation lasted more than 5 minutes it was converted to a regular rhythm by a brief reinfusion of cold organ preservation solution and this event recorded. Twenty minutes after reperfusion, biventricular reworking mode was restarted and measurements of hemodynamic performance were assessed at 10, 30, and 60 minutes of reworking mode. Notation was made whether the hearts were showing signs of ongoing improvement of aortic output and developed pressure before sacrifice (> 25% improved since previous measurement), and were then freeze clamped. Eight Collins, eight UW, and five Plegisol hearts served as nonreperfused controls and were freeze clamped immediately after the 3-hour ischemic interval.
Statistical analysis
Data were analyzed with BMDP software (Berkeley, California). Values were expressed as mean ± standard error of the mean. Differences between continuous variables were tested using matched Student’s t tests. Chi-square analyses were used for categorical values. Differences among groups were tested using ANOVA. Probability values less than 0.05 were considered significant.
|
Results
|
|---|
The average cannulation time (23 ± 1 minute), percent organs acceptable for experimentation (80%), and cannulae leak (7.1 ± 0.4 mL/min) were similar to that reported in the preliminary work with this model [3]. The hemodynamic values of right and left heart function before and after ischemia as well as the number of samples for each group, are presented in Tables 1 and 2, respectively. The common measured and derived parameters of hemodynamic function, and certain load-insensitive measurements were similar among the groups. Most items were significantly worse after ischemia except for some load-sensitive values, including peak systolic pressure and aortic output for Plegisol and several load-insensitive items for the Roe solution. It should be noted that the data for the UMC solution are favorably biased because seven hearts died in this group, preventing inclusion of even lower measurements to the group average. The organ preservation flow rates for the solutions were similar except for the more viscous UW solution (in milliliter per minute): UMC, 14.0 ± 0.9; Collins, 11.2 ± 0.6; Plegisol, 13.9 ± 0.6; UW, 6.9 ± 0.2; Custodiol, 15.1 ± 0.8; and Roe, 11.5 ± 0.7 (p < 0.001).
Several categorical observations could be made regarding the postreperfusion recovery of the hearts. The hearts generally survived, but higher mortality was seen in the UMC group: UMC, 33%; UW, 6.25%; Collins, 0%; Plegisol, 0%; Custodiol, 0%; and Roe, 0% (p = 0.003). Fibrillation was seen less in the Plegisol group: Plegisol, 21%; Collins, 24%; UW, 31%; UMC, 48%; Roe, 80%; and Custodiol, 100% (p = 0.004). Although many hearts (primarily from E solutions groups) had reached a plateau in their recovery at 30 minutes after reperfusion, other hearts (primarily from I solutions groups) continued to gain objective function at the 60-minute interval. Hearts receiving I solutions like Roe had more ongoing improvement at the time of sacrifice: Roe, 50%; Custodiol, 50%; Collins, 42%; UW, 6.25%; UMC, 0%; and Plegisol, 0% (p < 0.001).
Percent recoveries of preischemic values are also shown in Tables 1 and 2, and the groups were compared to each other. The parameters commonly used to measure postischemic recovery in the Langendorff heart preparation (ie, aortic output, developed pressure) favored the Plegisol group. In contrast, however, load-insensitive measurements like preload-recruitable stroke work and peak loop work area favored the Roe and Custodiol groups. Right ventricular recovery was similar among solutions. ATP measurements did not parallel hemodynamic performance. The postischemia reperfusion values for ATP (in micromoles per gram of protein) were UMC, 42.6 ± 2.4; Collins, 46.3 ± 6.3; Plegisol, 33.3 ± 4.2; and UW, 55.2 ± 8.1 (p = 0.14). The postischemia, nonreperfusion values for ATP (in micromoles per gram of protein) were Collins (n = 8), 70.6 ± 11.3; Plegisol (n = 5), 53.8 ± 8.2; and UW (n = 8), 75.9 ± 14.9 (p = 0.52).
|
Comment
|
|---|
The results of this study suggest that there are important differences between various heart organ preservation solutions in use today that are not obvious by clinical observation alone. In this study the ischemic times are relatively brief, typical for heart transplants that occur mostly within the home organ procurement areas. Except for one custom solution, the postischemic hemodynamic function was similar enough among groups to make recognition of variation unlikely by transplant surgeons.
Comparisons among solutions using more sophisticated measurements were more revealing. The data for UMC suggest that some custom solutions used for heart organ preservation may be inferior. The UMC solution was lactated Ringer’s based and the results in this model were not surprising given the adverse effect that lactate has on anaerobic metabolism. Nevertheless, this solution was applied to the preliminary cardiac transplant experience at our institution because it was used with clinical success for routine open heart operations at UMC. It should be emphasized that a solution that is not deleterious in a multidose cardioplegia environment may have different effects when used as a single infusion before prolonged ischemia. This is why many organ preservation solutions have greater acid-buffering capabilities than the bicarbonate used in Plegisol (a relatively poor buffer at low temperatures). The load-sensitive indices of recovery like aortic output flow and systolic pressure were quite good for Plegisol. This was also expected because these traditional values observed in the Langendorff preparation were used to optimize the components of Plegisol during its development. However, it is quite interesting that the load-insensitive measurements favored the I solutions like Roe. This brings into question whether some load-sensitive data from similar preparations could be misleading as it relates to postischemic recovery. As suggested by several other studies including our own, the efficacy of UW for short-term heart organ preservation may be impaired by its relatively high potassium concentration [13, 14]. The mechanism behind this possible deleterious effect is the opening of calcium channels during ischemia that is caused by hyperkalemia.
The findings in these experiments also correlate with emerging paradigms in clinical myocardial preservation. Although experiments in myocardial preservation typically focus on rapid recovery after ischemia, cardiac surgeons prefer to observe gradual recovery after cross-clamp removal. Rapid or hyperdynamic function after reperfusion may signify rapid calcium influx caused by cell membrane dysfunction or compensatory hypercontractility caused by the presence of endothelins or other mechanisms [15, 16]. The E solutions in this study showed an earlier plateau in function rather than the gradual recovery and better preload-recruitable stroke work and peak loop work reserve shown by Roe and the other I solutions. The UW did not behave like the other I solutions in this study and this same effect and reasons for it were reported in our United Network for Organ Sharing (UNOS) registry investigation (Table 3) [2]. It is difficult to know whether ventricular fibrillation that occurs during early reperfusion indicates a problem with myocardial preservation. It may be that electrical instability from heterogeneous delivery of reperfusate with differential rewarming is transient and has little to do with problems with myocardial energy metabolism.
The model used in these experiments offers a practical means to screen and compare various organ preservation solutions. Since our last description of this model, we found that preload-recruitable stroke work was more consistent than time-varying elastance as a load-insensitive parameter. This is attributed to the tendency of hearts to dilate after ischemia, causing diastolic "creep" of the length measurements (see Fig 2). This tends to increase the slope of the line that connects the end-systolic points. We also found that elastance measurements differed depending on whether the hearts were volume loaded or off-loaded during data acquisition. Furthermore, we found that ATP measurements were somewhat unreliable after ischemia as evidenced by high variability and lack of consistency with observed hemodynamic response. We did observe a decrease in ATP values after 60 minutes of reperfusion as compared with end-ischemia and it is also interesting that Plegisol had somewhat lower mean ATP levels.
We have demonstrated considerable variability in the practice of myocardial preservation both for routine open heart operations as well as cardiac transplantation and have attempted to cite possible causes for this [1, 2]. Hypothermia and mechanical cardiac arrest are such dominant protective factors that suboptimal solution components are unlikely to surface as noticeable problems for most cases. Also, technical speed and a variety of other intraoperative and patient factors affect outcome as well. Nevertheless, we know that myocardial preservation failures still occur and may exist at a subclinical level. The risk of such occult injury is high for allograft hearts subjected to prolonged ischemia because of ample functional reserve that can "hide" faulty preservation methods. It is also likely that such subtle differences are invisible unless the population of patients is quite large. This makes discovery of such problems, even at a "busy" transplant center, unlikely. Prospective study by large registries like UNOS, therefore, seems appropriate.
This study is potentially limited by the choice of animal used in the model, although it has been a research standard. Also, the lack of blood elements in this model will limit reperfusion injury effects. The organs used in this experiment were stored in balanced cold buffer solution to simulate the clinical scenario rather than the organ preservation solution. However, because of the cannulation method, the buffer bathed the epicardium primarily and could not enter the cardiac chambers and coronary system directly. Also, UW hearts received less of this viscous solution because infusion pressure was fixed; however, there was ample solution flow to attain tissue equilibrium for all hearts. There is the possibility of some confounding factors such as time or the later introduction of some preservation solutions; however, we made no changes in the apparatus and the methods used in this study were well tested in preliminary experiments. There were no differences in experimental yield/animal or other quality control measures followed during the course of the study. This experiment could have been improved by better correlative data than ATP measurements to link the load-insensitive hemodynamic observations with a mechanism of enhanced myocardial preservation. Also, these findings could be different with longer ischemic intervals and further study toward this end is in progress.
In summary, the results of this study support the preferred use of I solutions for heart organ preservation. Roe solution and emerging proprietary solutions like Custodiol are acceptable options. One should avoid the use of custom solutions unless tested experimentally, as it is difficult for individual practitioners to detect suboptimal solutions clinically (confounding issues aforementioned). Rather, large-scale prospective study of solution use and associated survival seems appropriate to confirm these findings and to achieve consensus regarding heart preservation solution usage.
|
Acknowledgments
|
|---|
This research is supported in part by a grant from the American Heart Association and the Harry S. Truman Memorial Veteran’s Hospital, Columbia, MO.
|
References
|
|---|
-
Demmy T.L., Haggerty S.P., Boley T.M., Curtis J.J. Lack of cardioplegia uniformity in clinical myocardial preservation. Ann Thorac Surg 1994;57:648-651.[Abstract]
-
Demmy T.L., Biddle J.S., Bennett L.E., Walls J.T., Schmaltz R.A., Curtis J.J. Organ preservation solutions in heart transplantation—patterns of usage and related survival. Transplantation 1997;63:262-269.[Medline]
-
Demmy T.L., Curtis J.J., Kao R., Schmaltz R.A., Walls J.T. Load-insensitive measurements from an isolated perfused biventricular working rat heart. J Biomed Sci 1997;1:111-119.
-
Demmy T.L., Magovern G.J., Kao R.L. Isolated biventricular working rat heart preparation. Ann Thorac Surg 1992;54:915-920.[Abstract]
-
Rannels D.E., Kao R., Morgan H.E. Effect of insulin on protein turnover in heart muscle. J Biol Chem 1975;250:1694-1701.[Abstract/Free Full Text]
-
Kao R.L., Trumble D.R., Magovern J.A., Christlieb I.Y., Magovern G.J. Fatigue resistant muscle with preserved force and mass for cardiac assist. J Card Surg 1991;6:210-217.[Medline]
-
Smith P.K., Krohn R.I., Hermanson G.T., et al. Measurement of protein using bicinchoninic acid. Anal Biochem 1985;150:76-85.[Medline]
-
Roe B.B., Hutchinson J.C., Fishman N.H., Ullyot D.J., Smith D.L. Myocardial protection with cold, ischemic, potassium-induced cardioplegia. J Thorac Cardiovasc Surg 1977;73:366-374.[Abstract]
-
Jeevanandam V., Auteri J.S., Sanchez J.A., et al. Improved heart preservation with University of Wisconsin solution: experimental and preliminary human experience. Circulation 1991;84:III324-III328.
-
Ledingham S.J., Braimbridge M.V., Hearse D.J. The St. Thomas’ Hospital cardioplegic solution. A comparison of the efficacy of two formulations. J Thorac Cardiovasc Surg 1987;93:240-246.[Abstract]
-
Toshima Y., Matsuzaki K., Mitani A., et al. The myocardial recovery mode after cold storage for transplantation with Collins’ solution and cardioplegic solution. A functional and metabolic study in the rat heart. J Thorac Cardiovasc Surg 1992;104:1320-1328.[Abstract]
-
Reichenspurner H., Russ C., Uberfuhr P., et al. Myocardial preservation using HTK solution for heart transplantation. A multicenter study. Eur J Cardiothorac Surg 1993;7:414-419.[Abstract]
-
Rosenfeldt F.L., Conyers R.A., Jablonski P., et al. Comparison of UW solution and St. Thomas’ solution in the rat: importance of potassium concentration. Ann Thorac Surg 1996;61:576-584.[Abstract/Free Full Text]
-
Amrani M., Ledingham S., Jayakumar J., et al. Detrimental effects of temperature on the efficacy of the University of Wisconsin solution when used for cardioplegia at moderate hypothermia. Comparison with the St. Thomas Hospital solution at 4 degrees C and 20 degrees C. Circulation 1992;86:II280-II288.
-
Kelly R.F., Hursey T.L., Schaer G.L., et al. Cardiac endothelin release and infarct size, myocardial blood flow, and ventricular function in canine infarction and reperfusion. J Investig Med 1996;44:575-582.[Medline]
-
Beyersdorf F., Acar C., Buckberg G.D., et al. Studies on prolonged acute regional ischemia. V. Metabolic support of remote myocardium during left ventricular power failure. J Thorac Cardiovasc Surg 1989;98:567-579.[Abstract]
Top
Abstract
Introduction
