TUESDAY MORNING, APRIL 25, 1995

7:00 am FORUM SESSION I

Auditorium, Hynes Convention Center

Moderators: Andrew S. Wechsler, M.D.

Douglas J. Mathisen, M.D.

F1. INHALED NO FAILS TO CONFER THE PULMONARY PROTECTION PROVIDED BY DISTAL STIMULATION OF THE NO PATHWAY AT THE LEVEL OF CGMP

Yoshifumi Naka, M.D., Ph.D.*, Dillip K. Roy, M.D.*, Arthur J. Smerling, M.D.*, Robert E. Michler, M.D.*, Craig R. Smith, M.D., David M. Stern, M.D.*, David J. Pinsky, M.D.* and †Mehmet C. Oz, M.D.*

New York, New York

Inhalation of nitric oxide (NO) gas effectively lowers pulmonary vascular resistance (PVR) in the adult respiratory distress syndrome, and some have suggested that it may be beneficial following lung transplantation. However, during reperfusion, NO combines rapidly with superoxide to form highly toxic peroxynitrite. We hypothesized that distal stimulation of the NO/cGMP pathway at the level of cGMP might confer the beneficial vascular effects of NO without its potential toxicities. To test this hypothesis, we used a recently described orthotopic rat left lung transplant model in which hemodynamic and survival measurements can be obtained independent of the native lung. Three conditions were tested: (1) Control (EC preservation solution), (2) Inhaled NO (EC preservation solution but inhaled NO given to recipient), and (3) EC preservation solution supplemented with the membrane permeable, nonhydrolyzable cGMP analog [8-(4-chlorophenylthio)-guanosine-3',5'-cyclic monophosphate; cGMP, 250 |iM]. The left lung was harvested from 22 male Lewis rats, preserved for 6 hrs at 4° C, and transplanted into isogenic recipients using cuff technique for all anastomoses. All recipients were ventilated with 70% ©2 and 30% N2 (as the carrier gas for NO was N2), which was supplemented with NO (65 ppm, monitored by chemiluminescence) in the second group. Gas mixtures were begun immediately prior to cross-clamp release and continued throughout reperfusion. In all experiments, the native (right) PA was ligated immediately after placement of a PA flow probe and Millar catheters into the LA and PA. PA flow (mL/min), arterial oxygenation (pO2, mm Hg) graft neutrophil infiltration (myeloperoxidase activity; MPO, A absorbance/min at 460 nm) and recipient survival (%) at 30 min. were measured. PVR (WU x 1000) was calculated. (Means ± SEM are shown; *=p<0.05 vs control, **=p<0.01 vs control; t=p<0.05 vsNO).

Only the cGMP analog was associated with reduced PVR, improved pO2, reduced graft neutrophil infiltration, and improved recipient survival. These data suggest that stimulation of the NO/cGMP pathway at the level of cGMP has a protective effect that is not seen with inhaled NO in the immediate pulmonary reperfusion period.

†Robert E. Gross Research Scholar

*By invitation

F2. DIRECT EFFECTS OF 3,5,3' TRIIODO-L-THYRONINE (T₃) ON MYOCYTE CONTRACTILE PROCESSES: INSIGHTS INTO MECHANISMS OF ACTION

‡Jennifer D. Walker, M.D.*, Rupak Mukherjee, M.S.*, Fred A. Crawford, M.D. and Francis G. Spinale, M.D., Ph.D.*

Charleston, South Carolina

Administration of 3,5,3' triiodo-L-thyronine (T₃) in the setting of congestive heart failure and cardiopulmonary bypass has been suggested to improve left ventricular (LV) function; however, the cellular basis for this improvement remains unknown. Accordingly, the present study examined the direct effects of T₃ administration on myocyte function and the contributory mechanisms for these effects. In isolated porcine LV myocytes (n=81), velocity of shortening increased in the presence of 80 pM T₃ compared to baseline (77.1±3.3 vs 117.0±5.0 jam/sec, p<0.05). In a separate series of experiments (n=29), pre-treatment with 80 pM T₃ increased P-adrenergic response (pAR: 25 nM isoproterenol) compared to PAR alone (203.7±16.2 vs 274.3±16.9 |im/sec, p<0.05). In isolated myocyte preparations (n=9), cyclic-AMP production (fmols/myocyte) was measured. Cyclic-AMP was unchanged with T₃, but T₃ pretreatment with PAR increased cyclic-AMP compared to PAR alone.

Since cyclic-AMP modulates intracellular Ca²⁺ processes, L-type Ca²⁺ processes, L-type Ca²⁺ channel density (patch clamp methods; pA/pF, n=15) and peak Ca²⁺ release from the sarcoplasmic reticulum (Fura-2 ionic measurement, pM, n=46) were measured. T₃ pretreatment followed by PAR increased L-type Ca²⁺ channel current and intracellular Ca²⁺ compared to PAR alone. The unique findings from this study are twofold. First, T₃ stimulates steady state myocyte contractile function by a cyclic-AMP independent mechanism. Second, T₃ potentiates the effects of PAR by increasing cyclic-AMP and subsequent phosphorylation of L-type Ca²⁺ channels. Thus, T₃ enhanced PAR transduction with resultant increased Ca²⁺ availability and improved myocyte inotropic response. These findings suggest the T₃ may provide a unique adjunct to conventional P-adrenergic agonist therapy in the clinical setting.

‡Nina S. Braunwald Research Fellow

*By Invitation

F3. INFLUENCE OF VEIN VALVES IN THE DEVELOPMENT OF ARTERIOSCLEROSIS IN RABBIT VENO-ARTERIAL GRAFTS

Aurelio Chaux, M.D.*, Xin Min Ruan, M.D.*, Michael C. Fishbein, M.D.* and Jack M. Matloff, M.D.

Los Angeles, California

Purpose: To study the influence of vein valves and define possible mechanisms for the formation of arteriosclerotic lesions in veno-arterial grafts.

Methods: Jugular vein grafts were interposed into the carotid circulations of 48 hypercholesterolemic rabbits. In 24 animals (Group A), vein segments did not contain vein valves; in another 24 (Group B), a vein valve was present. Subgroups of 6 animals were sacrificed at 2, 4, 6 and 8 weeks. Two mm sections were fixed in 4% paraformaldehyde and placed in sucrose for morphology and morphometry or frozen for immunohistochemistry and in situ hybridization for growth factor expression. Morphometry was done with computer-based image measuring software (Optimas) and validated by manual planimetry in 70% of animals. Cell identification and couting was done by immunohistochemistry. Dependent variables: graft mean area, mean thickness and proliferated cell nuclear antigen. Independent variables: presence or absence of vein valves and time in weeks post surgery. Measurements of mean area and thickness at cross sections were made just before and after the valve in Group B. Post valve values were also compared with those in the middle of the graft of Group A animals. Paired t-test and regression analysis were used.

Results: At 2 weeks intimal and medial thickening were due to an increase in cell number. From 2 to 6 weeks further intimal and medial thickening occurred without additional increase in cell number. At 6 weeks, foam cells and lipid deposits appeared, and at 8 weeks changes identical to human arteriosclerotic plaques were evident. Changes developed sooner and more intensely in Group B animals (p<001), and segments immediately after the valve developed them faster and more severely than before the valve (p<.001). PCNA staining showed greater cell proliferation in the post-valve segments (p<.001). Endothelial and intimal growth factor expression was strong for IGF-1 and EOF, mild for PDGF.

Conclusions: The presence of vein valves augments and accelerates intimal hyperplasia and arteriosclerosis, and at least 3 growth factors may be involved in the development of these lesions. To our knowledge, this is the first controlled experiment that demonstrates this phenomenon.

*By invitation

F4. MYOCARDIAL STUNNING: A THERAPEUTIC CONUNDRUM

Zhandong Zhou, M.D.*, Robert D. Lasley, Ph.D.*, Rolf Bunger, M.D, Ph.D.*, Julia Hegge, B.S.* and Robert M. Mentzer, Jr., M.D. Madison, Wisconsin and Bethesda, Maryland

Dobutamine (DOB) and pyruvate (PYR) are two potent inotropic agents with different mechanisms of action. While both can be used to treat post-ischemic myocardial dysfunction, the potential deleterious effects of augmenting myocardial contractility in the setting of myocardial stunning has not been addressed. To test the hypothesis that these agents may have adverse effects on the stunned heart, systolic wall thickening (SWT), myocardial phosphorylation potential index (CrP/CrxPj), interstitial fluid adenosine (ISF [ADO]) and myocardial oxygen consumption (MV02) were measured in open chest pigs. SWT was measured using sonomicrometry, and microdialysis probes were inserted to estimate ISF [ADO]. Stunning was induced with a 10 min occlusion of the anterior descending coronary artery. After 30 min reperfusion, pigs were treated with either DOB (10 jig/kg/min) or PYR (1 ml/min or a 150 mM solution). Results expressed as mean±SEM; *p<0.05 vs control, фp<0.05 vs stunned.

DOB produced myocardial de-energization, as the CrP/ CrxP_i. ratio decreased from 0.17±0.02 to 0.09±0.02 (p<0.02) in the stunned hearts. In contrast, PYR enhanced myocardial energy status, since the ratio increased from 0.20±0.03 to 0.55±0.08(p<0.01).

Infusion of both inotropic agents resulted in a marked improvement in regional SWT. The dobutamine effect, however, produced a marked increase in MVO₂ associated with an expected rising ISF [ADO] due to the decline in the CrP/CrxP_i. This energetic stress may account for the further deterioration in cardiac function observed after cessation of DOB. In contrast, PYR improved postischemic contractile function and enhanced myocardial energetics. These experimental findings suggest that under certain circumstances the clinical use of ß receptor agonists to treat myocardial stunning after cardiac surgery or heart transplantation may be suboptimal, if not undesirable. Further investigation is warranted to determine the optimum threapy for the stunned heart.

*By invitation

F5. PRECONDITIONING WITH POTASSIUM CHANNEL OPENERS: A NEW CONCEPT FOR ENHANCING CARDIOPLEGIC PROTECTION

Philippe Menasche, M.D., Ph.D., Egidijus Kevelaitis, Ph.D.*, Christian Grousset, M.D., Ph.D.*, Armand Piwnica, M.D. and Gerard Bloch, M.D.*

Paris, France

Background. Preconditioning (PC) is now recognized as one of the most effective means of improving the tolerance of myocardium to a sustained episode of ischemia. It has been suggested that this protection could result from an adenosine-induced opening of adenosine triphosphate-sensitive potassium channels and the related limitation of calcium overload.

Objective. This study was therefore designed to assess whether the protective effects of ischemic PC can be mimicked by the administration of a potassium channel opener in the surgically relevant setting of global myocardial ischemia.

Methods. Forty isolated isovolumic rat hearts were subjected to 45 minutes of normothermic potassium arrest followed by 1 hour of reperfusion. Hearts were randomly divided into four equal groups that only differed by the modalities of the prearrest treatment protocol: group 1 had no prearrest intervention (controls); group 2 hearts were preconditioned with 5 minutes of global ischemia followed by 5 minutes of buffer reperfusion prior to the onset of potassium arrest; in group 3, the preconditioning stimulus consisted of a 5-minute infusion of the potassium channel opener nicorandil (lOpM) followed by 5 minutes of drug-free buffer perfusion prior to potassium arrest; in group 4, the same protocol as in group 3 was used except that the administration of nicorandil was preceded by a 5-minute infusion of the selective potassium channel blocker glibenclamide ([glib]10|aM). During arrest, the left ventricular balloon was kept inflated to monitor the time course of ischemic contracture.

Results. They are summarized below (data analysis by 2-way ANOVA with repeated measures).

Pretreatment with nicorandil also lowered peak contracture during arrest significantly (40.4 ± 2.1 mmHg vs 51.8 ± 3.0 mmHg in controls [P<0.01 vs nicorandil], 50.0 ± 2.8 mmHg in ischemically preconditioned hearts [P<0.01 vs nicorandil] and 54.9 ± 5.3 mmHg in nicorandil+glibenclamide-treated hearts [P<0.01 vs nicorandil]) and lengthened the time to peak contracture, as compared with the control and nicorandil+glibenclamide groups (2,066 ± 45 sec vs 1,806 ± 91 sec [P<0.03] and 1,573 ± 104 sec [P<0.001], respectively).

Conclusion. The protective effects of ischemic PC can be mimicked by a pharmacological opening of adenosine triphosphate-sensitive potassium channels. The surgical relevance of these data stems from (1) the possibility to plan the onset of aortic cross-clamping and, consequently, to implement a PC stimulus in a timely appropriate fashion, and (2) the approval of nicorandil for clinical use in patients with coronary artery disease.

*By invitation

F6. THE ROLE OF NITRIC OXIDE IN CEREBRAL RECOVERY AFTER CARDIOPULMONARY BYPASS WITH DEEP HYPOTHERMIC CIRCULATORY ARREST IN INFANT PIGLETS

Takeshi Hiramatsu, M.D.*, Takuya Miura, M.D.*, Adre Duplessis, M.D.*, Masahiro Tanji, M.D.*, Joseph M. Forbess, M.D.*, Miles K. Tsuji, M.D.*, David Holtzman, M.D.* and Richard A. Jonas, M.D.

Boston, Massachusetts

Although nitric oxide (NO) is neurotoxic in vitro, there is continuing controversy regarding the role of NO in global cerebral ischemia. The effects on brain metabolic recovery of L-arginine (L-arg), a NO precursor, and L-Nitro-Arginine Methyl Ester (L-NAME), a NO synthase inhibitor, have been studied in an infant piglet model of deep hypothermic circulatory arrest (DHCA).

METHODS: Forty 2-week old piglets underwent core cooling to 15°C, 1 hour of DHCA at 15°C, 45 minutes of rewarming and 3 hours of normo-thermic reperfusion. Group L-arg (n=10) received 30 mg/kg of L-arginine IV before cardiopulmonary bypass (CPB) and 10 mg/kg/min infusion during the first 1 hour of reperfusion (Rep). Group L-NAME 10 mg/kg (n=10) received 10 mg/kg of L-NAME IV before CPB. Group L-NAME+L-arg (n=10) received both L-NAME and L-arg as in the first two groups. Control (n=10) received no medication. Recoveries of cerebral high energy phosphates and pHi were assessed by magnetic resonance spectroscopy in half of the animals in each group. Cerebral blood flow (CBF) and cerebral vascular resistance (CVR) by microspheres, cerebral metabolic rates of oxygen and glucose, and the redox state of cytochrome a,a3 by near infrared spectroscopy were assessed in the rest. Brain water content was measured in all animals after the experiments.

RESULTS: Results are given as mean±SEM. *p<0.05 vs. Control by Student-Newman-Keuls test

SUMMARY: L-arginine improved recovery of cerebral ATP, cyto-chrome a,a3 oxidation and cerebral blood flow. L-NAME reduced recovery of cerebral ATP, pHi and cytochrome a,a3 oxidation and increased mean blood pressure, CVR and brain water content. These effects of L-NAME were minimally reversed by L-arginine infusion after reperfusion.

CONCLUSIONS: The net effect of intra-neuronal and endothelial nitric oxide is to promote cerebral metabolic recovery after global ischemia.

*By invitation

F7. THE EFFECTS OF RETROGRADE CEREBRAL PERFUSION AFTER PARTICULATE EMBOLIZATION TO THE BRAIN

Mustafa E. Yerlioglu, M.D.*, Craig K. Mezrow, M.S.*, All M. Sadeghi, M.D.*, Peter S. Midulla, M.D.*, Ning Zhang, M.D.*, Howard H. Shaing, D.V.M.*, Donald J. Weisz, Ph.D.* and Randall B. Griepp, M.D.

New York, New York

Neurologic injury as a consequence of cerebral embolism of air or atherosclerotic debris during cardiac and aortic surgery remains a major cause of morbidity. We have developed a chronic porcine model to evaluate the potential of retrograde cerebral perfusion (RCP) for improving cerebral outcome following cardiac surgery, and have previously demonstrated that it results in a small amount of nutritive flow, and provides cerebral protection superior to prolonged hypothermic circulatory arrest. We designed the current study to evaluate the efficacy of RCP in mitigating the effects of cerebral embolism of paniculate matter during cardiac surgery.

Four groups of pigs (20-28 kg) were assigned to undergo deep hypothermia at an esophageal temperature of 20°C: a retrograde control group (RC, n=5); a retrograde embolism group (RE, n=10); an antegrade control group (AC, n=5); and an antegrade embolism group (AE, n=6). The sagittal sinus (SS) was cannulated in all animals, and flow in the SVC was regulated in order to achieve and maintain a SS pressure of 30 mmHg during RCP. During ante-grade perfusion, the aortic arch pressure was maintained at 50 mmHg. Polystyrene microspheres, (250-750 um, 200 mg), were utilized as the embolic material. In the embolism groups, the spheres were injected via a cannula in the isolated arotic arch; controls were injected with 10 cc of saline. Five minutes after injection, antegrade perfusion was continued in antegrade groups, and retrograde perfusion via an SVC cannula was instituted in the retrograde groups, both for a duration of 25 minutes. Blood returning to the aortic arch was collected and measured in the RCP groups. All animals were allowed to recover for at least five days and were evaluated daily using a quantitative behavioral score, in which 9 indicates complete normalcy, 7 means that the animals were able to stand unassisted and were likely to recover fully, and lower numbers indicate substantial injury, with 0=death. At the time of elective sacrifice, half of the brain was utilized for recovery of embolized spheres after digestion with ION NaOH.

As shown in the table, neurological recovery in the control groups was complete with retrograde as well as with antegrade perfusion. Following embolization, there was severe neurological injury in both groups, and both antegrade and retrograde flows were reduced compared with controls. Although fewer spheres were present in the brains of the RE group, neurologic outcome was no better in the group as a whole then in the AE group. When the RE group was examined in greater detail, however, it was seen that some animals recovered almost completely after retrograde perfusion (score>7, good outcome). Further scrutiny showed that these good outcome RE animals required significantly lower SVC pressure to achieve adequate SS pressures than the animals that sustained severe neurologic injury following embolization and RCP. Although the mechanism of the injury following use of high SVC pressures during RCP after embolization (but not in controls) remains unclear, our data suggest that RCP during cardiac surgery may mitigate cerebral injury from participate emboli if adequate RCP can be achieved using SVC pressure <40 mmHg.

F8. EFFECTS OF CARDIOPLEGIA ON VASCULAR FUNCTION AND THE NO-REFLOW PHENOMENON FOLLOWING ISCHEMIA-REPERFUSION: STUDIES IN THE ISOLATED BLOOD-PERFUSED RAT HEART

Vincenzo Argano, M.D., FRCS*, Manuel Galinanes, M.D., Ph.D.*, Stephen J. Edmondson, MRCP, FRCS* and David J. Hearse, Ph.D., D.Sc.*

London, United Kingdom

Sponsored by: Mark V. Braimbridge, London, United Kingdom

The protective role of cardioplegia against post-ischemic cardiac dysfunction is well established. However, the effect of cardioplegia on vascular function and its repercussion on the no-reflow phenomenon are controversial. We have investigated the influence of St. Thomas' cardioplegic solution on: (i) endo-thelium-dependent and endothelium-independent vascular function (EDVF and EIVF), and (ii) the extent of the no-reflow phenomenon. Isolated rat hearts (n=16/group) perfused with blood at 60 rnmHg, were subjected to 10, 20, 30 or 40 min of global ischemia and 40 min of reperfusion (37°C). Eight hearts in each group also received cardioplegia (40 mmHg for 2 min) before ischemia. At the end of reperfusion, a bolus of 250ug (3mM) of nitro-L-arginine methyl ester was infused to assess EDVF. After 20 min washout, 25ug (0.3mM) of sodium nitroprusside was infused to assess EIVF. Fluorescein (1 ml, 1% w/v) was then infused to assess no-reflow. Hearts were frozen, cut into transverse sections (lOxlmm), video-recorded under UV light, and images digitised and analysed for density of fluorescence. No-reflow was defined as a flow <5% (corresponding to 171-256 grey-scale density). EIVF (percent of non-ischemic control) was severely decreased only after 30 min and 40 min of ischemia (45. ± .6% and 31. ± 2.5%; p<0.05), but was significantly protected by cardioplegia (62. ± .7% and 57. ± .9%). A significant reduction in EDVF was only observed after 40 min of ischemia (69. ± 0.6%; p<0.05) and again this was improved by cardioplegia (89. ± .8%). Areas of no-reflow were present after 30 and 40 min of ischemia (11. ± .8% and 33. ± 4.1% of LV mass) and they were significantly decreased by cardioplegia (0. ± .4% and 3. ± .6%; p<0.05). In conclusion, cardioplegia protects against post-ischemic endothelium-dependent and endothelium-independent vascular dysfunction and reduces the extent of no-reflow.

*By invitation

9:00 am PLENARY SESSION

Auditorium, Hynes Convention Center

Moderators: Robert B. Wallace, M.D.

James L. Cox, M.D.

15. SURGICAL MANAGEMENT OF EXTENSIVE

CALCIFICATION OF THE MITRAL VALVE ANNULUS

Alain F. Carpentier, M.D., Ph.D., Jean-François Fuzelier, M.D.*, John Y.M. Relland, M.D.* and Michel Pellerin, M.D.*

Paris, France

Extensive calcification of the mitral valve annulus (MVA) is a pathological entity frequently associated with degenerative valvular disease. The calcification process involves at least 1/3 of the circumference of the MVA and usually not the valvular tissue. It may extend however to the underlying myocardium. Whenever an operation is necessary for an associated valve insufficiency, the question arises whether it is preferable to repair or to replace the valve and how to manage the calcification. This paper reports the technical details, the recent improvement and the mid-term results of a reconstructive operation proposed in 1986 which up to now has only been the object of early and brief reports on small series. This operation comprises the temporary detachment of the corresponding leaflets, en bloc resection of the calcium deposit with its surrounding fibrous capsula, annular reconstruction and valve repair. For patients in whom the calcification process extends to the myocardium, a modified operation has been more recently developed which comprises in addition a "sliding plasty" of the left atrium over the area of resected calcium.

Between 1986 and 1994, among 63 patients (pts) with extensive calcification of the annulus and severe mitral valve insufficiency, 62 benefited from these techniques of repair. Ages ranged from 18 to 81 years (mean 62). Twenty-eight pts had a billowing mitral valve (Barlow), 32, a fibroelastic deficiency and 2, Marfan's disease. The calcification involved 1/3 of the annulus in 26 pts, 2/3 in 34 pts and the whole annulus in 2 pts. In 8 pts, the calcification process extended to the papillary muscles (4 pts) and/or deeply within the myocardial wall (6 pts). There were 2 hospital deaths (3.2%) and no early reoperation. The follow-up period extended from 4 months to 8 years (mean 3 years 8 months). There were 2 late deaths, 2 and 17 months after the operation (one valve related) for an actuarial survival of 95.7% at 7 years.

Late reoperation (3 to 62 months) was necessary in 4 pts (6.4%) for residual mitral valve incompetence (MVI) (n=2), hemolysis (n=1) or endocarditis (n=1). In one of these pts, a new repair was possible, whereas the 3 other pts required a valve replacement. All patients but one survived the reoperation. Actuarial freedom from reoperation was 88% at 7 years.

All 56 patients with valve repair were reviewed for this study by clinical examination and echocardiography. All but one were in functional chiss I or II. There was no MVI in 19, trivial MVI in 17 and moderate MVI in 7. No thromboembolic complications have been recorded.

This study shows that complete annulus decalcification and valve repair can be carried out safely in patients with mitral valve insufficiency and associated extensive calcification of the annulus, even when the calcification process deeply involves the myocardium. It also demonstrates that an initially good result remained stable up to 7 years.

*By invitation

16. POSTINFARCTION VENTRICULAR SEPTAL RUPTURE: REPAIR BY ENDOCARDIAL PATCH WITH INFARCT EXCLUSION

Tirone E. David, M.D., Laura Dale, R.N.*, Robert J. Cusimano, M.D.* and Zhao Sun, B.A.*

Toronto, Ontario, Canada

Because the extent of right ventricular infarction has been identified as an important determinant in the outcome of pts with postinfarction ventricular septal defect (VSD), an operation that avoids further damage to the right ventricle was developed and it has been applied in 43 pts during the past 8 years. The procedure is performed by opening the left ventricle through the infarcted wall and by suturing a pericardial patch (fresh autologous or bovine pericardium) to the endocardium of the left ventricle with a continuous 3-0 polypropylene suture to exclude the infarcted muscle from the left ventricular cavity. The ventriculotomy is closed primarily and no infarctectomy is performed.

There were 21 men and 22 women, ages 48 to 84 years, mean 68. Twenty-seven pts were in cardiogenic shock when operated on and all had an intraaortic balloon pump inserted preoperatively; 7 of them also required assisted ventilation and 13 were anuric. Doppler echocardiography was used to confirm the diagnosis and assess left and right ventricular function in all pts. The VSD was anterior in 22 pts and posterior in 21. All pts had coronary angiography. Pts in shock were operated on as an emergency and most pts who were hemodynamically stable were operated on within 3 days of diagnosis. In addition to the endocardial pericardial patch, 29 pts had coronary artery bypass and one pt also had mitral valve replacement. There were 6 operative deaths (14%); 3 in pts with anterior VSD and 3 in pts with posterior VSD. Eleven pts required renal dialysis and 15 required assisted ventilation for more than 48 hours. ICU stay (mean ± SD) was 7.8 ± 10.3 days and hospital stay was 20 ± 18 days. Only one pt had a recurrent VSD but it closed spontaneously within one month. Operative survivors have been followed from 3 to 95 months, mean of 42. There were 7 late deaths, 4 cardiac. The actuarial survival at 7 years was 64% ± 8%. The only two factors that were predictive of mortality were renal failure and left ventricular ejection fraction <35%, by a stepwise logistic regression analysis. All pts are in NYHA class I or II. Doppler echocardiography was performed at least once a year in all survivors and revealed mild impairment of left ventricular function in 18 pts, moderate in 11 and severe in 1. All pts have normal or near normal right ventricular function.

This new operative technique is relatively simple and appears to improve the outcome of pts with postinfarction VSD.

*By invitation

17. AORTIC VALVE REPAIR IN ACUTE TYPE A DISSECTION: IS IT SOUND?

Ludwig K. von Segesser, M.D.*, Urs Niederhauser, M.D.*, Marietta Schonbeck, M.D.*, Paul Vogt, M.D.* and Marko I. Turina, M.D.

Zurich, Switzerland

The last 200 consecutive patients with acute Stanford type A dissection (157 men [78%], 43 women [22%]) were analyzed in order to assess the validity of aortic valve repair whenever possible. Indication for surgery was in most cases based on echocardiographic examination only in order to reduce the doctors' delay. In the majority of patients (111/200: 56%), the incompetent aortic valve was resuspended and repaired. Aortic root replacement with a compositd graft was performed in 66/200 patients (33%) mainly because of enlarged aortic anulus and sinus portion. Replacement of the aortic valve and the supracoronary ascending aorta was performed in 23/200 patients (12%) with diseased aortic valve (e.g., bicuspid valve) but acceptable aortic sinus portion. Follow-up totalized 656 patient years (maximum 14 years). Actuarial analyses as a function of aortic valve type provided the following probabilities ± errors (95%).

During long term follow-up, there was no significant difference between groups with regard to structural deterioration, valve thrombosis, thrombo-embolic complications, anticoagulation-induced hemorrhage, and endocarditis. Freedom from valve failure and valve-related complications are similar for repaired, mechanical, and biological valves. Consequently, valve-related reoperations are rare during at least five years of follow-up. Hence, aortic valve repair in acute type A dissection can be recommended in the majority of patients.

10:05 am INTERMISSION - VISIT EXHIBITS

*By invitation

10:50 am PLENARY SESSION

Auditorium, Hynes Convention Center

Moderators: Robert B. Wallace, M.D.

James L. Cox, M.D.

18. PREDICTORS OF SURVIVAL IN MALIGNANT TUMORS OF THE STERNUM

Nael Martini, M.D., Andrew G. Huvos, M.D.*, Michael E. Hurt, M.D., Ph.D., Robert Heelan, M.D.*, Manjit S. Bains, M.D., Patricia M. McCormack, M.D., Valerie W. Rusch, M.D., Michael Weber, S.A.C.* and Robert J. Ginsberg, M.D.

New York, New York

From 1930 to 1994, 54 patients with primary malignant tumors of the sternum were seen. The median age was 54 years (range: 19-78 years). Fifty patients presented with a mass, one-half of them also had pain in the sternal region. Two patients had no symptoms at presentation. Of 39 solid tumors, 26 were chondrosarcomas, 10 osteosarcomas, 1 fibrosarcoma, 1 angiosarcoma and 1 malignant fibrous histiocytoma. Of these, 25 were low grade and 14 were high grade tumors. Of 15 small cell tumors, 8 were plasmacytomas, 6 malignant lymphomas, and 1 Ewing's sarcoma.

The size of the tumor ranged from 3 to 24 cm. in maximum diameter (median 7 cm.). Twenty-one patients had only plain chest roentgenograms, 26 had CT scans, 15 plain tomograms and 7 had MR scans. Tissue diagnosis was established by needle aspiration in 3 patients, open biopsy in 38 and at surgery in 13.

Partial or subtotal sternectomy was done in 36 patients and total sternectomy in 3 (34/39 sarcomas and 5/15 small cell tumors). Of the remaining 15 patients, 4 had local excision, 10 external radiation and/or chemotherapy and one had no treatment. All but one patient treated by wide resection (N=39) had some form of skeletal reconstruction of the chest wall defect. Twenty-three were repaired with marlex mesh and methylmethacrylate, 7 with marlex mesh, 3 with ox fascia, 2 with fascia lata, 1 with rub struts + fascia lata, 1 with tentalum mesh, and 1 by shifting pectoral muscles to cover the defect. The skin edges were closed by primum in 37 patients; 2 required skin grafts.

There were 2 deaths secondary to respiratory failure (one due to inadequate stabilization following total sternectomy and one due to pneumonia). Two infections necessitated removal of the methylmethacrylate prosthesis.

In solid tumors, adverse prognostic factors were high grade tumor, pulmonary metastases, inadequate or no resection and local recurrence.

Resection in chondrosarcomas yielded a 5-year survival (Kaplan-Meier) of 80% (median follow-up: 17 years). The 5-year survival in osteosarcomas was 14%. Resection was curative in 54% of low grade sarcomas but in only 7% of high grade sarcomas. The local recurrence rate following complete resection was 14%. In small cell tumors, resection and radiation were helpful for local control; all failures were due to distant metastases.

We conclude that primary sarcomas of the sternum though uncommon are potentially curable by wide surgical excision. With rigid prostheses to repair the skeletal defects, the surgical complication rates are minimized and survival following treatment remains tumor dependent.

*By invitation

19. TRANSMYOCARDIAL LASER REVASCULARIZATION: INITIAL CLINICAL EXPERIENCE

Denton A. Cooley, M.D., O.H. Frazier, M.D., Kamuran A. Kadipasaoglu, Ph.D.*, Seckin Pehlivanoglu, M.D.*, Eddy Barasch, M.D.*, Matthias Lindenmeir, M.D.*, Dena P. Houchin, R.N.*, Wanda Samuels, R.N.*, K. Lance Gould, M.D.* and Susan Wilansky, M.D.*

Houston, Texas

Transmyocardial laser revascularization (TMLR) with an 850-W CO₂ laser involves drilling 1-mm diameter channels into a beating heart after left thoracotomy. Clotting acutely occludes the channels on the subepicardium (SEp) and, in the long term, camerosinusoidal connections improve subendo-cardial (SEn) perfusion. We performed TMLR in 21 consecutive patients (pts; mean age, 63 y; 3 women) who had hibernating myocardium and/or reduced coronary flow reserve by positron emission tomography (PET) and low-dose dobutamine echocardiography (DE). All pts had distal diffuse coronary artery disease refractory to antianginal therapy. The mean angina class (AC_m, Canadian Cardiovascular Society) was 3.76 (4 pts with unstable angina) and the mean resting left ventricular ejection fraction (LVEF) by MUGA was 47%.

During follow-up, AC_m was 2.40 at 3 mo (n=15 pts, p<0.01), and 1.5 at 6 mo (n=l 1 pts, p<0.001). METs (ml Oj/kg/min) on treadmill improved from 3.2 at baseline (BL) to 4.4 at 3 mo (p=NS), and to 5.6 at 6 mo (p<0.05). Resting LVEF increased but not significantly at either time. On PET analysis at 3 mo, the mean SEn/SEp perfusion ratio increased by 8% in lased regions but decreased by 4% in nonlased septal regions (n=10 patients, p<0.01). On DE, the percentage of normokinetic segments (15-segment analysis) had increased from 46% at BL to 57% at 3 mo (p=NS) and to 62% at 6 mo (p<0.05). Of 7 (33%) patients who had an adverse event after TMLR (5 deaths, 2 re-revascularizations), 4 (57%) had a history of anterior/anterolateral myocardial infarction (vs 0% of 14 non-event pts, p<0.001), 6 (86%) were being treated for congestive heart failure (vs 36%, p<0.001), and 5 (71%) had regurgitant mitral valve disease (vs 29%, p<0.05). Histopathologic examination of myocardium from one of the patients who died revealed multiple patent, endothelium-lined neochannels connected to the native sinusoidal vasculature.

These results suggest that, in patients with reversible myocardial ischemia, the laser channels remain functional in the long term and augment subendo-cardial perfusion. We conclude that TMLR produces immediate clinical benefit and improves cardiac function at 6 mo.

*11:30 am ADDRESS BY HONORED SPEAKER

Medical Ethics in the 21st Century: DNR or CPR?

Edmund D. Pellegrino, M.D., Washington, D.C.

12:10 pm ADJOURN FOR LUNCH - IN EXHIBIT HALL

12:10 pm CARDIOTHORACIC RESIDENTS' LUNCHEON Independence Ballroom, Sheraton Boston Hotel