Ann Thorac Surg 1998;65:1260-1264
© 1998 The Society of Thoracic Surgeons
Myocardial Oxygenation During Terminal Warm Blood Cardioplegia
Michio Kawasuji, MDa,
Shigeyuki Tomita, MDa,
Tamotsu Yasuda, MDa,
Naoki Sakakibara, MDa,
Hirofumi Takemura, MDa,
Yoh Watanabe, MDa
a Department of Surgery (I), Kanazawa University School of Medicine, Kanazawa, Japan
Accepted for publication December 6, 1997.
Address reprint requests to Dr Kawasuji, Department of Surgery (I), Kanazawa University School of Medicine, Takaramachi 13-1, Kanazawa 920, Japan
e-mail: (kawasuji{at}med.kanazawa-u.ac.jp)
 |
Abstract
|
|---|
Background. Terminal warm blood cardioplegia accelerates myocardial metabolic recovery. The process of myocardial oxygenation during terminal warm blood cardioplegia and its optimal administration are not clear.
Methods. We measured the myocardial tissue oxygen saturation (SO2) during reperfusion using near-infrared spectroscopy. Twenty-four dogs underwent 1 hour of ischemic arrest with cold crystalloid cardioplegia. They were then divided into four equal groups. Group 1 dogs received normal blood reperfusion. The other dogs received 15 mL/kg of terminal warm blood cardioplegia at 80 mm Hg in group 2 or at 60 mm Hg in group 3, and 30 mL/kg of cardioplegia at 60 mm Hg in group 4, followed by blood reperfusion.
Results. In group 1, the SO2 increased gradually during the early reperfusion and decreased transiently during the late reperfusion. In group 2, the SO2 increased rapidly but it decreased transiently during blood reperfusion. In groups 3 and 4, the SO2 increased rapidly and remained at high levels during the blood reperfusion. Reperfusion ventricular fibrillation occurred along with a SO2 decrease only in groups 1 and 2. The postischemic troponin-T levels of groups 3 and 4 were lower than that of group 1. The functional recovery in group 4 was better than those in the other three groups.
Conclusions. Terminal warm blood cardioplegia accelerates the early SO2 increase and abolishes the SO2 decrease during subsequent reperfusion and reduces the incidence of reperfusion arrhythmia, suggesting that it ameliorates reperfusion injury and consequently improves postischemic functional recovery.
|
Introduction
|
|---|
Reperfusion-induced injury is a paradoxic extension of ischemic damage, which occurs during reperfusion after myocardial ischemia [1]. The consequences of reperfusion injury include reperfusion arrhythmias and myocardial stunning (ie, the prolonged contractile dysfunction that occurs despite absence of irreversible injury) [2–4]. There is evidence suggesting a link between the two major mechanisms of reperfusion injury—impaired calcium homeostasis and the formation of oxygen free radicals during early reperfusion [3, 4]. The net result is an increased transsarcolemmal calcium influx and cellular calcium overload. Metabolically, reperfusion injury is characterized by a reduced capacity of the heart to take up and use oxygen [1, 5, 6]. Terminal warm blood cardioplegia is administered to lower oxygen demands by keeping the heart in an arrested state during the initial phase of reperfusion, when the postischemic oxygen utilization capacity is impaired most, and to allow the heart to channel its energy resources toward reestablishing ionic and cellular homeostasis, while optimizing the metabolic rate of repair with normothermia [6–9]. However, the myocardial oxygen metabolism during terminal warm blood cardioplegia has not been fully clarified, because monitoring devices to allow the determination of myocardial oxygenation in the arrested heart have not been developed.
Near-infrared (NIR) light passes through biological tissue with relative ease and is significantly absorbed by oxygenated and deoxygenated hemoglobin, which have distinctly different absorption spectra in the NIR region [10]. On the basis of this difference, changes in the tissue concentration of each hemoglobin can be measured by NIR spectroscopy [11–13]. Three-wavelength NIR spectroscopy has been developed to measure tissue oxygen saturation and the tissue hemoglobin concentration [14, 15]. This apparatus allows the real-time monitoring of changes in myocardial tissue oxygenation in the cardioplegically arrested heart.
The present study was undertaken to examine changes in myocardial tissue oxygenation during reperfusion with normal blood or warm blood cardioplegia after hypothermic cardioplegic arrest, and to determine the optimal administration of terminal warm blood cardioplegia.
|
Material and methods
|
|---|
Twenty-four adult mongrel dogs weighing 12 to 17 kg were studied. The animals were anesthetized with an intramuscular administration of ketamine hydrochloride (20 mg/kg) and intravenous sodium pentobarbiturate (30 mg/kg) and were mechanically ventilated with a volume respirator. The dogs received humane care in compliance with the "Principles of Laboratory Animal Care" (National Society for Medical Research) and the "Guide for the Care and Use of Laboratory Animals" published by the National Institutes of Health (NIH publication 85-23, revised 1985). A cardiopulmonary bypass was performed with the use of a membrane oxygenator and a centrifugal pump. During the cardiopulmonary bypass, the flow rate was kept at 80 mL/kg per minute; the mean arterial pressure was kept at 60 mm Hg and the systemic temperature was maintained at 32°C. After the cross-clamp of the aorta, 15 mL/kg of cold crystalloid cardioplegic solution was administered into the aortic root. One liter of cardioplegic solution contained 30 mEq of potassium, 24 g of glucose, 7 g of mannitol, and 1.2 mL of 7% sodium bicarbonate for a pH of 7.5. The dogs received 10 mL/kg of cold cardioplegic solution every 20 minutes during a 60-minute period of aortic cross-clamp. After the 60-minute ischemic arrest period, dogs were divided into four groups of 6 dogs. In group 1, after adequate rewarming, the aortic cross-clamp was released and the heart received normal warm (37°C) blood reperfusion at an aortic root pressure of 60 mm Hg. In groups 2, 3, and 4, before the release of the aortic cross-clamp, warm (37°C) blood cardioplegia solution was infused through an aortic root using a cardioplegia infusion set (BCD Advanced system; Shiley, Irvine, CA). The dogs received 15 mL/kg of warm blood cardioplegia for 1 minute at 80 mm Hg in group 2 or for 2.5 minutes at 60 mm Hg in group 3, or received 30 mL/kg of warm blood cardioplegia for 5 minutes at 60 mm Hg in group 4 (Fig 1). The blood cardioplegia consisted of blood mixed 4:1 with crystalloid cardioplegic solution and an additional dose of potassium chloride, yielding a concentration of 15 mEq/L. After the aortic clamp release, the heart resumed beating or was defibrillated, if necessary. Each dog underwent an additional 45 minutes of total vented cardiopulmonary bypass before extracorporeal circulation was discontinued.
View larger version (27K):
[in this window]
[in a new window]
|
Fig 1. Schematic presentation of the experimental protocol. (CCC = cold crystalloid cardioplegia; TWBC = terminal warm blood cardioplegia.)
|
|
Myocardial tissue oxygenation was measured using a three-wavelength NIR spectroscopy apparatus (PSA-IIIN; Biomedical Science, Kanazawa, Japan). The principle of this technique is as follows. The iron–porphyrin moieties of oxygenated and deoxygenated hemoglobin have distinctly different absorption spectra at NIR wavelengths (700 to 1,000 nm) that readily penetrate intact tissue [10]. The relationship between the concentration of a chromophore and the intensity of light transmitted through a solution is expressed by the Beer-Lambert law. When the light is transmitted in a homogeneous tissue, the tissue concentration of oxygenated hemoglobin and deoxygenated hemoglobin can be measured in accordance with this law [11–14]. Because oxygenated hemoglobin and oxygenated myoglobin each have essentially identical NIR absorption spectra, hemoglobin plus myoglobin are added together. The NIR spectroscopy probe (PSP-15R; Biomedical Science) was attached to the anterior surface of the right ventricle, avoiding epicardial fat. This probe contains three light-emitting diodes as light sources and three pairs of silicone photodiodes to detect the intensity of reflected light [15]. The light-emitting diodes delivered NIR light at the three wavelengths (700, 730, 770 nm). The distance between the light-emitting diode and the photodiode and the distance between the paired photodiodes were each set at 2.5 mm. Thus, optical information from a myocardial tissue depth of 2.5 to 5.0 mm was obtained. The NIR signal was analyzed using a set of algorithms written to solve for oxygenated and deoxygenated hemoglobin plus myoglobin. Oxygenated hemoglobin plus myoglobin divided by the sum of oxygenated and deoxygenated hemoglobin plus myoglobin corresponded to the tissue oxygen saturation (SO2). Because most hemoglobin is located in capacitant vessels, changes in tissue SO2 are most sensitive to changes in the capillary and venous oxygen content.
Hemodynamic data were recorded before and after cardiopulmonary bypass using an arterial catheter, a left atrial catheter, and a Swan-Ganz thermodilution catheter (American Edwards Laboratory, Santa Ana, CA). The cardiac index (CI) in mL · min-1 · kg-1 and stroke work index (SWI) in g · m/kg were calculated by the following equations: 
, where CO is cardiac output in mL/min, MAP is mean aortic pressure, LAP is mean left atrial pressure, HR is heart rate, and BW is body weight in kilograms.
The serum levels of troponin-T were measured by an enzyme-linked immunosorbent assay to assess myocardial injury during the procedure. Samples were examined before and 3 hours after the cardiopulmonary bypass.
Cumulative data are expressed as the mean ± the standard deviation of the mean. Statistical analysis was performed with Student’s t test and analysis of variance and Scheffé’s F test to detect significant (p < 0.05) differences between measured variables.
|
Results
|
|---|
The baseline value of myocardial SO2 during beating was 81% ± 2% in the four groups. The myocardial SO2 decreased to 71% ± 3% at induced ventricular fibrillation and decreased further to 57% ± 4% at the infusion of crystalloid cardioplegic solution (Fig 2). In group 1, the myocardial SO2 increased gradually to 81% ± 1% at the initial phase of blood reperfusion after the release of the aortic cross-clamp, but it then decreased gradually to 72% ± 3% (p < 0.01 versus baseline) 2.5 minutes after the start of the reperfusion. This decrease in myocardial SO2 was associated with the occurrence of ventricular fibrillation for all 6 dogs. After electrical defibrillation, the myocardial SO2 returned gradually to the baseline level. In groups 2, 3, and 4, the myocardial SO2 increased rapidly during the initial phase of the administration of terminal warm blood cardioplegia (Table 1). The hearts were asystolic during the administration of terminal warm blood cardioplegia, which took approximately 1, 2.5, and 5 minutes in groups 2, 3, and 4, respectively. In group 2, the myocardial SO2 increased to 86% ± 3% at the infusion of warm blood cardioplegic solution, but it rapidly decreased to 75% ± 5% (p < 0.01 versus baseline) after the release of the aortic clamp, and ventricular fibrillation occurred concomitantly in 4 dogs, in which the myocardial SO2 decreased to less than 75%. In groups 3 and 4, the myocardial SO2 increased to 82% ± 1% at the administration of warm blood cardioplegic solution. The hearts began to beat spontaneously 30 seconds after the release of the aortic clamp and did not show any significant decrease in myocardial SO2 for all 12 dogs. There were no significant differences in myocardial SO2 between groups 3 and 4 throughout the procedure.
View larger version (28K):
[in this window]
[in a new window]
|
Fig 2. Changes in the myocardial tissue oxygen saturation of the dogs. (A = during beating; B = infusion of cold crystalloid cardioplegia; C = normal blood reperfusion or infusion of warm blood cardioplegia; SO2 = myocardial tissue oxygen saturation; TWBC = terminal warm blood cardioplegia; VF = ventricular fibrillation.)
|
|
There were no differences in hemodynamic parameters measured before the cardiopulmonary bypass among the four groups (Table 2). The postischemic cardiac performance was decreased in all four groups. The functional recovery after reperfusion was better in groups 3 and 4 than in groups 1 and 2. The postischemic cardiac performance in group 4 was better than that in group 3, although the difference between the two groups did not reach significance.
Serum troponin-T levels before cardiopulmonary bypass were less than 0.1 ng/mL. Troponin-T levels were markedly increased after cardiopulmonary bypass in all four groups. Troponin-T levels of groups 3 and 4 were significantly low compared with that of group 1, although there was no significant difference between these two groups (Fig 3).
View larger version (27K):
[in this window]
[in a new window]
|
Fig 3. Serum troponin-T levels before and after cardiopulmonary bypass in the dogs. Postischemic troponin-T levels of groups 3 and 4 were significantly (p < 0.05) lower than that of group 1.
|
|
|
Comment
|
|---|
The major findings of this study are as follows. (1) The hearts subjected to normal blood reperfusion after hypothermic cardioplegic arrest showed a characteristic change in myocardial SO2, a slow increase at the initial phase followed by a transient decrease during the subsequent phase of reperfusion. (2) The administration of terminal warm blood cardioplegia enhanced the initial increase in myocardial SO2 and abolished its transient decrease, resulting in a low incidence of reperfusion ventricular fibrillation and in improved postischemic ventricular performance.
The present study focused on the process of myocardial reoxygenation during reperfusion after cardioplegic arrest; although previous studies suggest that hypocalcemia, alkalosis, and substrate enrichment are effective to enhance cardioplegia protection [6, 8]; the effects of the cardioplegic component were not examined. The present canine study showed that the rates of increase in the myocardial SO2 during the initial phase of reperfusion with normal blood and warm blood cardioplegia differed significantly. The myocardial SO2 increased gradually at normal blood reperfusion. On the other hand, the myocardial SO2 increased rapidly at the administration of terminal warm blood cardioplegia. Digerness and colleagues [16] reported that coronary flow was less and coronary vascular resistance higher during the initial phase of blood reperfusion after cold crystalloid cardioplegia in humans. Ferguson and associates [17] reported that coronary vascular regulation remained abnormal within 3 minutes after reperfusion after hypothermic blood cardioplegic arrest. The impairment of the coronary vasodilation reserve is most likely caused by impaired endothelium-dependent microvascular reactivity, which is observed after ischemic arrest with cold crystalloid cardioplegia [18, 19]. It is possible that terminal warm blood cardioplegia maintains low coronary vascular resistance and enhances oxygen delivery to the myocardial tissue.
The major metabolic deficit caused by ischemic myocardial damage is a limited capacity to utilize delivered oxygen [6, 8]. Kane and colleagues [5] found that the mitochondrial oxygen uptake remains reduced after reperfusion, and identified a defect in electron transport in the respiratory chain. In the present study, the myocardial SO2 returned to the preischemic baseline level, but it then showed a significant decrease 2.5 minutes after start of blood reperfusion. This decrease in myocardial SO2 is considered to be attributable to an increase in metabolic requirements as postischemic electromechanical activity became forceful [20]. A previous study of postcardioplegia reperfusion showed that global myocardial blood flow was decreased at 3 and 6 minutes after reperfusion and that myocardial oxygen consumption increased as a result of electromechanical activity [21]. In the present study, the hearts subjected to terminal warm blood cardioplegia maintained a high level of myocardial SO2, which did not show any decrease at the subsequent blood reperfusion. These results suggest that terminal warm blood cardioplegia may enhance myocardial reoxygenation and optimize the oxygen supply/demand balance of the myocardium during reperfusion.
Reperfusion-induced arrhythmia is partly a consequence of reperfusion injury after ischemia, and may range in severity from ventricular premature beats to ventricular fibrillation. Our experimental study showed that ventricular fibrillation occurred in all 6 dogs subjected to normal blood reperfusion and in 2 of the 6 dogs subjected to a 1-minute period of terminal warm blood cardioplegia, when the myocardial SO2 decreased to below 75% during blood reperfusion. The decrease in myocardial SO2 during reperfusion suggests a negative oxygen supply/demand balance. Ventricular fibrillation per se may reduce the myocardial SO2 further as myocardial oxygen consumption increases. We observed that the myocardial SO2 returned to a baseline level only after defibrillation. Conversely, the hearts subjected to adequate terminal warm blood cardioplegia did not show any decrease in myocardial SO2 at the subsequent blood reperfusion or showed no incidence of reperfusion ventricular fibrillation. Previous studies reported that terminal warm blood cardioplegia reduced the incidence of reperfusion ventricular fibrillation after 45 minutes of normothermic ischemia and even after 8 hours of ischemic preservation [22, 23].
Myocardial stunning after cardioplegic arrest is another important consequence of reperfusion injury after ischemia [2, 3]. As in previous reports [6–9], the present study found that the hearts that received terminal warm blood cardioplegia showed a better postischemic functional recovery than the hearts that received normal blood reperfusion. We compared three techniques of administration of terminal warm blood cardioplegia. The 1-minute period of terminal warm blood cardioplegia was not effective for the recovery of myocardial oxygen metabolism, as evidenced by a significant decrease in the myocardial SO2 associated with a high incidence of ventricular fibrillation during subsequent blood reperfusion. Conversely, the 2.5- and 5-minute periods of terminal warm blood cardioplegia seem effective for the recovery of myocardial oxygen metabolism. The postischemic functional recovery after the 5-minute period of terminal warm blood cardioplegia was better than those after blood reperfusion and the 1- and 2.5-minute periods of blood cardioplegia. These results were supported by the findings that the myocardial injury assessed by troponin-T after the 2.5- and 5-minute periods of terminal warm blood cardioplegia was less than that after the normal blood reperfusion. The duration of its administration is an important factor of terminal warm blood cardioplegia [20, 24]. Studies in hearts reperfused with warm blood cardioplegia showed that the oxygen uptake was augmented during the first 2 minutes and then declined during the next 7 to 9 minutes [22]. A prolonged period of terminal warm blood cardioplegia is disadvantageous in that it prolongs the cardiopulmonary bypass time and increases the potassium load.
One limitation of NIR spectroscopy is that myocardial oxygenation data are regional. It should be assumed that NIR spectroscopy readings are representative of global myocardial oxygenation.
In conclusion, reperfusion with normal blood after cold cardioplegic arrest resulted in abnormal coronary vascular regulation and reduced myocardial oxygenation, the process of which was monitored by changes in the myocardial SO2. Reperfusion ventricular fibrillation occurred along with a decrease in the myocardial SO2. These abnormalities were ameliorated by the use of terminal warm blood cardioplegia. NIR monitoring of myocardial SO2 is useful to achieve the optimal administration of terminal warm blood cardioplegia.
|
References
|
|---|
- Hearse D.J. Reperfusion of the ischemic myocardium. J Mol Cell Cardiol 1977;69:605-616.
- Braunwald E., Kloner R.A. The stunned myocardium: prolonged, postischemic ventricular dysfunction. Circulation 1982;66:1146-1149.[Abstract/Free Full Text]
- Bolli R. Mechanism of myocardial "stunning". Circulation 1990;82:723-738.[Abstract/Free Full Text]
- Hearse D.J. Myocardial injury during ischemia and reperfusion: concepts and controversies. In: Yellon D.M., Jennings R.B., eds. Myocardial protection. New York: Raven Press, 1992:13-34.
- Kane J.J., Murphy M.L., Bisset J.K., DeSoyza N., Doherty J.E., Staub K.D. Mitochondrial function, oxygen extraction, epicardial S-T segment changes and tritiated digoxin distribution after reperfusion of ischemic myocardium. Am J Cardiol 1975;36:218-224.[Medline]
- Rosenkranz E.R., Buckberg G.D. Myocardial protection during surgical coronary reperfusion. J Am Coll Cardiol 1983;1:1235-1246.[Medline]
- Follette D., Steed D., Foglia R., Fey K., Buckberg G.D. Reduction of postischemic myocardial damage by maintaining arrest during initial reperfusion. Surg Forum 1977;28:281-283.[Medline]
- Lazar H.L., Buckberg G.D., Manganaro A., Becker H., Mulder D.G., Maloney J.V., Jr Limitations imposed by hypothermia during recovery from ischemia. Surg Forum 1980;31:312-315.
- Teoh K.H., Christakis G.T., Weisel R.D. Accelerated myocardial metabolic recovery with terminal warm blood cardioplegia (hot shot). J Thorac Cardiovasc Surg 1986;91:888-895.[Abstract]
- Jöbsis F.F. Noninvasive infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 1977;198:1264-1267.[Abstract/Free Full Text]
- Hampson N.B., Piantadosi C.A. Near infrared monitoring of human skeletal muscle oxygenation during forearm ischemia. J Appl Physiol 1988;64:2449-2457.[Abstract/Free Full Text]
- Parsons W.J., Rembert J.C., Bauman R.P., Greenfield J.C., Jr, Pianntadosi C.A. Dynamic mechanisms of cardiac oxygenation during brief ischemia and reperfusion. Am J Physiol 1990;259(Heart Circ Physiol 28):H1477-H1485.
- Vaughan D.L., Wickramasinghe Y.A.B.D., Russell G.I., et al. Near infrared spectroscopy: blood and tissue oxygenation in renal ischemia-reperfusion injury in rats. Int J Angiol 1995;4:25-30.
- Narita N., Tominaga T., Koshu K., Mizoi K., Yoshimoto T. Monitoring of brain tissue hemoglobin concentration and oxygen saturation using a three wavelength spectrophotometric method. Neurol Res 1994;16:428-432.[Medline]
- Kawasuji M., Yasuda T., Tomita S., Sakakibara N., Takemura H., Watanabe Y. Near-infrared monitoring of myocardial oxygenation during intermittent warm blood cardioplegia. Eur J Cardio-thorac Surg 1997;12:236-241.[Abstract/Free Full Text]
- Digerness S.B., Kirklin J.W., Naftel D.C., Blackstone E.H., Kirklin J.K., Samuelson P.N. Coronary and systemic vascular resistance during reperfusion after global myocardial ischemia. Ann Thorac Surg 1988;46:447-454.[Abstract/Free Full Text]
- Ferguson E.R., Spruell R.D., Vicente W.V.A., Murrah C.P., Holman W.L. Coronary vascular regulation during postcardioplegia reperfusion. J Thorac Cardiovasc Surg 1996;112:1054-1063.[Abstract/Free Full Text]
- Sellke F.W., Shafique T., Schoen F.J., Weintraub R.M. Impaired endothelium-dependent coronary microvascular relaxation after cold potassium cardioplegia and reperfusion. J Thorac Cardiovasc Surg 1993;105:52-58.[Abstract]
- Murphy C.O., Pan-Chih, Gott J.P., Guyton R.A. Microvascular reactivity after crystalloid, cold blood, and warm blood cardioplegic arrest. Ann Thorac Surg 1995;60:1021-1027.[Abstract/Free Full Text]
- Allen B.S., Okamoto F., Buckberg G.D., Leaf J., Bugyi H. Studies of controlled reperfusion after ischemia. XII. Effects of "duration" of reperfusate administration versus reperfusate "dose" on regional functional, biochemical, and histochemical recovery. J Thorac Cardiovasc Surg 1986;92:594-604.[Abstract]
- Holman W.L., Vicente W.V.A., Spruell R.D., Digerness S.B., Pacifico A.O. Effect of postcardioplegia reperfusion rhythm on myocardial blood flow. Ann Thorac Surg 1994;58:351-358.[Abstract/Free Full Text]
- Kofsky E.R., Julia P.L., Buckberg G.D. Overdose reperfusion of blood cardioplegic solution. A preventable cause of postischemic myocardial depression. J Thorac Cardiovasc Surg 1991;101:275-283.[Abstract]
- Tian G.H., Mainwood G.W., Biro G.P., et al. The effect of high buffer cardioplegia and secondary cardioplegia on cardiac preservation and postischemic functional recovery. A 31P NMR and functional study in Langendorff perfusion pig hearts. Can J Physiol Pharmacol 1991;69:1760-1768.[Medline]
- Buckberg G.D. Myocardial temperature management during aortic clamping for cardiac surgery: protection, preoccupation, and perspective. J Thorac Cardiovasc Surg 1991;102:895-903.[Abstract]
This article has been cited by other articles:
|
|
|
|
 
C. Rergkliang, A. Chetpaophan, V. Chittithavorn, P. Vasinanukorn, and V. Chowchuvech
Terminal Warm Blood Cardioplegia in Mitral Valve Replacement: Prospective Study
Asian Cardiovascular and Thoracic Annals,
April 1, 2006;
14(2):
134 - 138.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. J. Penttila, M. V.K. Lepojarvi, K. T. Kiviluoma, P. K. Kaukoranta, I. E. Hassinen, and K. J. Peuhkurinen
Myocardial preservation during coronary surgery with and without cardiopulmonary bypass
Ann. Thorac. Surg.,
February 1, 2001;
71(2):
565 - 570.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Kawasuji, M. Ikeda, N. Sakakibara, S. Fujii, S. Tomita, and Y. Watanabe
Near-infrared monitoring of myocardial oxygenation during ischemic preconditioning
Ann. Thorac. Surg.,
June 1, 2000;
69(6):
1806 - 1810.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
