TUESDAY MORNING, APRIL 28, 1992

7:30 a.m. FORUM SESSION I - Cardiac Surgery - Los Angeles Ballroom

F1. Pathogenesis of Ischemic Mitral Insufficiency

MARIO R. LLANERAS, M.D.*,

STEPHEN W. DOWNING, M.D.*, JOAO A. C. LIMA, M.D.*,

MICHAEL L. NANCE, M.D.*, RADU C. DEAC, M.D.,

PHILIP L. LINDEN, B.A.* and

L. HENRY EDMUNDS, JR., M.D.

Philadelphia, Pennsylvania

We hypothesized that the combination of a moderate sized posterior in-farct and infarction of the posteromedial papillary muscle (PPM) were required to cause ischemic mitral regurgitation (MR); neither condition alone would cause MR. In 72 sheep (19 slaughterhosue hearts, 25 acute experiments and 28 chronic experiments) the coronary artery anatomy was detailed by gross inspection, dye injections in cadaveric hearts, vital staining and coronary angiography. Ligation of the first and second marginal (OM1 & OM2) branches of the left circumflex artery consistently infarcted 23 ± 3.3% (by planimetry) of the left ventricular mass. Color flow doppler performed in-traoperatively in 12 sheep with the OM1/OM2 infarct, revealed no appreciable MR up to eight weeks postinfarct despite dilatation on the left ventricle (LV).

Ligation of the second and third (OM2 & OM3) marginal branches, in 16 sheep, infarcted 21.4 ± 4.0% of the LV mass and completely infarcted the PPM in all but two animals. Wall thickness at the level of the PPM is 1.6 ± 0.2 cm before infarction and decreases to 0.4 ± 0.1 cm 8 weeks after infarction. Left ventriculography and intraoperative color flow echocardiography demonstrated MR which progressed in severity following infarction. Eight sheep have been followed serially for 8 weeks; two had incompletely infarcted PPM and did not develop IMR. Three other sheep developed severe mitral regurgitation immediately after infarction and died early. Five sheep with MR have not reached 8 weeks (cf. graph). The consistency of coronary arterial anatomy in sheep and lack of collaterals result in anatomically uniform infarcts between sheep and predictable ischemic MR. The data show that a 23% posterior infarction sparing the PPM or an infarction of the PPM alone does not produce MR. Both conditions must be present. This preparation provides a reproducible model of ischemic MR and offers a means to devise repairs based on the pathophysiology of the disease.

*By Invitation

F2. Very Small Diameter Polyurethane Vascular Prostheses With Rapid Endothelialization for Coronary Artery Bypass Grafting

TAKAFUMI OKOSHI, M.D.*, GIORGIO SOLDANI, Sc.D.*

MOSES GODDARD, M.D.* and

PIERRE M. GALLETTI, M.D.*

Providence, Rhode Island

Sponsored by: Karl E. Karlson, M.D.,

Providence, Rhode Island

The growing incidence of "redo" coronary artery bypass grafting (CABG) calls for the development of satisfactory prosthetic substitutes for the internal mammary artery or the saphenous vein. Porous non-woven tubular fabrics provide one approach to that challenge. Two types of spongy polyurethane-polydimethylsiloxane (Cardiothane 51™, Kontron Instruments, Inc.) vascular grafts with an internal diameter (ID) of 1.5 mm were fabricated by a spray, phase-inversion technique. Low porosity grafts (LPG) with hydraulic permeability (HP) of 2.7 ± 0.4 ml/min.cm², and medium porosity grafts (MPG) with HP of 39 ± 8 ml/min/cm², displayed good handling properties and suturability. Twelve straight LPG, seventeen straight MPG (1.5 - 2.0 cm in length) and one loop MPG (10 cm in length) were implanted by the same surgeon end-to-end in the infrarenal aorta of 30 male Sprague-Dawley rats. No antithrombogenic agents were administered pre- or postoperatively. At 3 months postimplantation, patency was 8% for LPG (1/12) and 76% for the straight MPG (13/17). The loop MPG was also patent. The sole patent LPG showed neointimal hyperplasia and incomplete endothelialization. All but one of the patent straight MPG showed a glistening and transparent neoin-tima with complete endothelialization. The loop MPG displayed endothelialization from each anastomosis and in many islands in the middle portion of the graft, totalling 47% of the luminal surface by morphometric analysis. Thick mural thrombus, anastomotic hyperplasia, or aneurysm formation were not observed in patent MPG. These data indicate that in the rat aortic replacement model it is possible to achieve patency and a high degree of endothelialization in very small diameter prostheses of appropriate porosity. This may open an avenue to the use of synthetic grafts in aorto-coronary bypass surgery.

*By Invitation

F3. Cardiopulmonary Dysfunction Produced by Initiating Reoxygenation on Cardiopulmonary Bypass in Immature Hypoxemic Piglets: Prevention by Intravenous Metabolic Treatment

GEORG MATHEIS, M.D.* GERALD D. BUCKBERG, M.D.,

DENIS B. TIXIER, M.D.*, HELEN H. YOUNG, Ph.D.*

and MICHAEL P. SHERMAN, M.D.*

Los Angeles, California

This study tests the hypothesis that reoxygenation injury is produced when Cardiopulmonary bypass (CPB) is initiated in immature hypoxemic piglets, and causes cardiopulmonary dysfunction that can be avoided by intravenous metabolic treatment before and during CPB.

Of eighteen immature Yorkshire-Duroc piglets (<3 weeks old) six were anesthetized, instrumented and observed over 5 hours (control). Twelve piglets underwent up to 2 hours of hypoxemia (paO₂ = 20-30 mmHg) before initiation of reoxygenation on CPB. Six received an intravenous metabolic infusate (mercaptopropionyl glycine, catalase, aspartate, glutamate, glucose/insulin) which was started before and continued during CPB.

Hypoxia produced an initial hyperdynamic response (39% increased cardiac index) followed by progressive hemodynamic deterioration. This required premature initiation of bypass in 8 out of 12 hypoxemic piglets (67%).

(values are mean ± SEM,* = p<0.05 vs. Control, ANOVA, ^arterial/alveolar pO, Ratio)

Reoxygenation induced injury (assessed 30 minutes after CPB) was characterized by reduced stroke work index, increased myocardial lipid peroxidation (conjugated dienes), increased pulmonary vascular resistance index, impaired statis lung compliance, and decreased a/A pd ratio. These reoxygenation changes were avoided by intravenous metabolic treatment.

We conclude that the reoxygenation of immature hypoxemic piglets by initiating CPB results in cardiopulmonary dysfunction which may increase vulnerability to subsequent ischemia (i.e. aortic crossclamping) and is preventable by intravenous metabolic treatment before and during CPB needed for cardiac repair.

*By Invitation

F4. Normocalcemic Blood or Crystalloid Cardioplegia Provides Superior Neonatal Myocardial Protection Over Low Calcium Cardioplegia

JEFFREY M. PEARL, M.D.*, HILLEL LAKS, M.D.,

DAVIS C. DRINKWATER, M.D.*,

AVEDIS MENESHIAN, B.S.* and PAULA. CHANG*

Los Angeles, California

Although standard blood cardioplegia provides good myocardial protection for adult cardiac surgery, protection of the cyanotic, immature myocardium remains suboptimal. Calcium, which has been implicated in reperfusion injury and the development of "stone heart" in mature myocardium, is routinely removed from standard cardioplegia solutions. Immature, neonatal myocardium has lower intracellular calcium stores and is more reliant on extracellular calcium for contraction, and for maintenance of the glycocalyx membrane. In order to determine if normocalcemic cardioplegia would result in improved cardiac function in the neonatal heart, we conducted a series of experiments using an isolated, blood perfused working heart model. Neonatal piglet hearts (24-48 hours) were excised without intervening ischemia and placed directly on a blood perfused circuit. Baseline stroke-work index (SWI) was assessed. Hearts were then arrested with cold cardioplegia delivered at 45 mmHg for 2 minutes; Group I = low-Ca + 2 Blood CP (Ca = 0.6 mmol/1); Group II = normal-Ca + 2 blood cardioplegia (Ca=1.1 mmol/1); Group III = University of Wisconsin (UW) solution; and Group IV = UW with added calcium (Ca= 1.0 mmol/1). Each group consisted of 8 hearts. Cardioplegia was administered every 20 minutes for 2 hours and topical hypothermia was employed. Hearts were then reperfused with warm whole blood. Functional recovery, expressed as percent of control SWI, was determined 60 minutes following reperfusion.

Hearts preserved with normocalcemic CP (Groups II and IV) had complete (100%) functional recovery at 60 minutes, whereas hearts preserved with low-calcium CP (Groups I and III) had only 80% and 58% recovery at LAP of 9 mmHg, respectively. Percent recovery decreased further at higher left atrial pressures in the low calcium groups, indicating diastolic dysfunction. Electronmicrographs taken 1 hour after reperfusion showed minimal edema and only mild myofibrillar changes, and were identical in both the low calcium and normal calcium groups. Conclusion: 100% functional recovery is possible in immature myocardium when calcium is added to either blood or crystalloid cardioplegia. The addition of calcium does not result in ultrastructural damage ("calcium paradox") and results in superior functional recovery.

By Invitation

F5. Effect of Calcium and Preischemic Hypothermia on Recovery of Myocardial Function After Cardioplegic Ischemia in Neonatal Lambs

MITSURU AOKI, M.D.*, JOHN E. MAYER, JR., M.D.,

FUMIKAZU NOMURA, M.D.* and

HIROAKI KAWATA, M.D.*

Boston, Massachusetts

Most neonatal cardiac operations are peformed at deep hypothermia, but controversy exists over the danger of hypothermia prior to cardioplegic arrest. Hypothermia can cause myocyte Ca accumulation experimentally and has been noted to result in reduced postischemic recovery in some experimental and clinical reports. Preischemic hypothermia is routinely used in other centers with good results. To explore this discrepancy we evaluated the effects of preischemic ionized (i) Ca concentration and temperature on outcome after 2 hrs and 15°C cardioplegia-protected ischemia in 38 isolated, blood perfused neonatal lamb hearts. LV maximal developed pressure (DP), dP/dt, -dP/dt, coronary blood flow (CBF), and oxygen consumption (MVO₂) were measured before and 30 and 60 minutes after 2 hours of ischemia. After baseline measurements, hearts were perfusion-cooled (groups B, C, and D) for 10 minutes to 17 °C and then arrested with cold (4°C) St. Thomas cardioplegia. Group A had 10 minutes of normothermic perfusion after baseline measurements, then were arrested with the cold cardioplegia. Group B had cooling normal iCa. Group C had citrate (Cit) added as cooling was started to low iCa and iCa was not normalized until 15 minutes into reperfusion. Group D received citrate plus Ca to give normal iCa during cooling. Results are given as % recovery of baseline except iCa (nM/1).* =p<0.05 vs groups B and D.

MVO₂ per beat significantly increased while the coronary blood flow decreased during preischemic cooling (p<0.05) in all groups. CBF was higher in Group C than in other groups during both pre and postischemia as long as iCa was low (p<0.05).

These data suggest that preischemic hypothermia results in reduced postischemic recovery of function than simultaneous induction of cardioplegia and hypothermia. Low iCa during preischemic hypothermia and early reperfusion offsets these effects. Therefore, careful attention to both temperature and iCa prior to ischemia are important in the outcome after ischemia in neonatal hearts.

*By Invitation

F6. Temporary Leukocyte Depletion Reduces Ventricular Dysfunction During Prolonged Postischemic Reperfusion

IAN C. WILSON, MB, ChB*, TIMOTHY J. GARDNER, M.D.,

JOSEPH M. DiNATALE, B.S.*,

A. MARC GILLINOV, M.D.*, WILLIAM E. CURTIS, M.D.*

and DUKE E. CAMERON, M.D.*

Baltimore, Maryland

Previous experiments in our laboratory have demonstrated that leukocyte depletion improves early postischemic ventricular performance in a neonatal model of global myocardial ischemia. However, the rate at which leukocytes return to the circulation after cardiopulmonary bypass (CPB), their possible late accumulation in the myocardium and their subsequent effect on the functional recovery of left ventricle is not known.

The present study examined the effect of leukocyte depletion on myocardial performance during the 6 hour period post CPB in an in situ, in vivo porcine model of neonatal cardiac surgery. Median sternotomy was performed on thirteen 3-5 day old piglets, 6 controls and 7 leukocyte depleted animals (LD), and left ventricular (LV) short-axis sonomicrometry crystals and an in-traventricular micromanometer positioned. Piglets were cooled to 22 °C prior to 90 minutes hypothermic ischemia after a single dose of cold crystalloid car-dioplegia. Mechanical leukocyte depletion was achieved using Pall RC100 filters. Granulocyte counts in the initial coronary reperfusate were reduced to 0.8% of controls (2±1 cells/ml, mean±SEM, p<0.001). However, circulating granulocyte counts progressively increased throughout the period of myocardial reperfusion reaching 66% of controls (874±356/ml) after 6 hours (p<0.3). Control piglets demonstrated an immediate reduction in postischemic left ventricular performance (measured by preload recruitable stroke work, PRSW) to 87 ± 6% baseline. PRSW further declined to a nadir of 73 ± 8% at 4 hours before improving to 86 ± 6% 6 hours postoperatively. In contrast, LV performance in LD animals were 94 ±6% baseline immediately post CPB and did not decline throughout the period of observation, remaining 98 ± 5% and 98 ± 4% baseline at 4 and 6 hours, respectively (p<0.02). Left ventricular systolic function (measured by end-systolic pressure-volume relationship) and ventricular compliance similarly were better preserved in the LD group throughout the postoperative period (p<0.05 and p<0.04, respectively). This improvement in postischecmic ventricular function was associated with decreased myocardial water content at 6 hours (79.6 ± 0.9%) compared to the control group (80.9 ± 0.8%, p<0.02), and with reduced tissue myeloperoxidase concentration (46% of controls,p<0.05), representing less myocardial leukocyte accumulation.

These data show that leukocyte depletion during initial reperfusion reduces myocardial leukocyte accumulation and results in sustained improvement of postischemic LV function. Temporary manipulation of the leukocyte population during cardiopulmonary bypass can reduce myocardial injury and improve myocardial protection following prolonged reperfusion despite rapid return of granulocytes to the circulation within 6 hours.

*By Invitation

F7. Alternative Methods of Retrograde Cardioplegia Delivery: Effects on Preservation of the Ischemic Left Ventricle After Acute Coronary Artery Occlusion and Reperfusion

JAMES T. DIEHL, M.D.* MICHAEL PONTORIERO, M.D.*,

RAYMOND CONNOLLY, Ph.D.*

STEVEN SCHWARTZ, M.D.* and

RICHARD J. CLEVELAND, M.D.

Boston, Massachusetts

Myocardial protection may be enhanced by altering the flow (intermittent vs continuous) and temperature (cold vs warm) of retrograde blood car-dioplegia. 24 dogs were randomized into 4 cardioplegia groups (n = 6): Group I - intermittent cold antegrade, Group II - intermittent cold retrograde, Group III - continuous warm retrograde, and Group IV - continuous cold retrograde. Preservation was assessed with a model of left ventricular ischemia and reperfusion induced by acute occlusion of the LAD. Functional parameters were collected at baseline, at 60 minutes of LAD occlusion (ischemia), and following reperfusion. All dogs were maintained normother-mic during cardiopulmonary bypass (CPB) and all data was collected off CPB. Regional and global LV function is reported as the load independent slope (E max) of fiber segment and minor LV axis dimension vs LV end systolic pressure respectively. Group I demonstrated no recovery of global LV function (#).Recovery of global LV function was similar for Grp II, III, and IV (*). Regional LV function demonstrated significant functional recovery only for Group IV, continuous cold retrograde cardioplegia (**). 2-D echo wall motion studies confirm this data.

Conclusions:

1. Retrograde cardioplegia provides superior global LV preservation during ischemia when compared with antegrade cardioplegia.

2. Preservation of regional myocardium beyond an occluded coronary artery is best achieved with cold continuous retrograde cardioplegia.

*By Invitation

F8. Detrimental Effects of Interrupting Warm Blood Cardioplegia During Coronary Revascularization

HIROSHI MATSUURA, M.D.*, HAROLD L. LAZAR, M.D.,

XI MING YANG, M.D.*, SAMUEL RIVERS, B.S.*,

PATRICK R. TRAENOR, CCP* and

RICHARD J. SHEMIN, M.D.

Boston, Massachusetts

Warm Blood Cardioplegia (WBC) has emerged as an alternative method of myocardial protection to Cold Blood Cardioplegia (CBC). However, the continuous infusion of blood required in this technique may obscure the operative field necessitating that the WBC be interrupted. The effects of interrupting WBC during coronary revascularization are unknown. This experimental study was therefore undertaken to determine whether interrupting WBC during coronary revascularization will result in increased myocardial damage.

In 30 adult pigs, the second and third diagonal vessels were occluded with snares just beyond the LAD for 1 ½ hours. All animals were then placed on cardiopulmonary bypass and underwent 45 minutes of cardioplegic arrest. Following aortic unclamping, the coronary snares were released and all hearts were reperfused for 3 hours. During the period of cardioplegic arrest, 10 pigs received intermittent, antegrade/retrograde CBC (4°C), 10 animals received continuous, retrograde WBC (37°C) at 100 ml/min; and 10 animals received continuous, retrograde WBC but had the infusion stopped for three 7 minute periods during the 45 minute crossclamp period. The effectiveness of myocardial protection in the area at risk was assessed by myocardial pH measured during cardioplegic arrest using tissue pH probes, Wall Motion Scores (WMS) using 2-D echo (4 = normal to -1 = dyskinesia), and the Area of Necrosis/Area of Risk (AN/AR) using histochemical staining. Results are Mean±SE;*p<.05 from Antegrade/Retrograde CBC; ₊p<.05 from Retrograde WBC.

Interrupting Retrograde WBC resulted in more tissue acidosis during acrioplegic arrest, lower wall motion scores, and increased tissue necrosis. We conclude that interrupting WBC during coronary revascularization diminishes the effectiveness of WBC and results in increased ischemic damage.

*By Invitation

F9. Complete Prevention of Myocardial Stunning, Low-Reflow and Edema After Heart Transplantation by Blocking Leukocyte Adhesion Molecule During Reperfusion

JOHN G. BYRNE, M.D.*, LAWRENCE H. COHN, M.D.,

WENDEL J. SMITH, M.D.*, MICHAEL P. MURPHY, M.D.*,

GREGORY S. COUPER, M.D.* and

ROBERT F. APPLEYARD, Ph.D.*

Boston, Massachusetts

Following heart preservation and transplantation, reperfusion-induced microvascular inflammation with release of leukocyte-derived toxic mediators (oxygen free radicals, enzymes) has been suggested as an underlying mechanism leading to edema, "low-reflow" and subsequent diastolic dysfunction ("diastolic stunning"). For lekocytes to mediate inflammation and release their toxic mediators they must first adhere to either each other or to the endothelium. We therefore hypothesized that preventing leukocyte adhesion, by blocking leukocyte membrane adhesion molecule CD18, would reduce myocardial inflammation and edema and improve reflow and diastolic function after heart preservation and transplantation.

Methods: After cardioplegia and insertion of an LV balloon, rabbit hearts were heterotopically transplanted into recipient rabbits either immediately (Immediate, n = 12) or after preservation in 4°C saline (3 hrs ischemia, n = 23). Recipients of preserved hearts received, 45 min before reperfusion, either IV saline (Placebo, n = 13) or IV anti-CD18 monoclonal antibody R15.7 (2 mg/Kg) (Anti-CD18, n =10). During 3 hours reperfusion the slope of the end-systolic pressure-volume relation (Emax), the exponential elastic coefficient of the end-diastolic pressure-volume relation (β), the unstressed ventricular volume (V₀) and the time constant of the exponential LV pressure fall after dP/dt_min (τ) were serially measured. Myocardial blood flow was measured with microspheres from which coronary vascular resistance (CVR) was calculated. After explantation the degree of myocardial inflammation, estimated by tissue leukocyte sequestration (Myeloperoxidase assay, MPO), and myocardial water content (%H₂O) were determined.

Values: Mean ± SEM;*p<0.01 vs Immediate; †p<0.01 vs Placebo; #p<0.05 vs Immediate; ‡p<0.05 vs Placebo (ANOVA)

Increased MPO, prolonged t, elevated CVR and increased %H₂O observed in Placebo hearts were all completely averted by Anti-CD18 treatment, demonstrating complete prevention of myocardial inflammation, diastolic stunning, low-reflow, and edema.

Conclusion: Leukocyte adhesion is a critical initiating event in reperfusion-induced myocardial damage after heart transplantation and should be addressed in treatments directed at limiting reperfusion injury.

*By Invitation

F10. A Long-Term Ventricular Assist System

WILLIAM S. PIERCE, M.D., ALAN J. SNYDER, Ph.D.*,

GERSON ROSENBERG, Ph.D.*, WILLIAM WEISS, M.S.*,

WALTER E. PAE, JR., M.D. and

JOHN A. WALDHAUSEN, M.D.

Hershey, Pennsylvania

While cardiac transplantation is excellent therapy for certain patients with end-stage heart disease, the number of donor hearts is far less than the need. A permanent left ventricle-to-aortic assist pump has the potential of providing long-term systemic circulatory support in patients for whom donor hearts cannot be obtained.

Our multidisciplinary group is developing an implantable, electrically powered pump that will provide tether-free circulatory support. The blood pump consists of a seamless polyurethane sac within a polysulfone case. Bjork-Shiley monostrut valves provide unidirectional flow. The blood sac is compressed by a pusher plate actuated by a brushless DC electric motor-motion translator.

The system has gradually evolved from one in which a tube crossed the skin, serving as a conduit for electrical wires, to the most recent model in which the unit is completely sealed and inductive coupling is used to transfer electrical energy across the intact skin. The current model has an implantable miniature control system as well as a battery that provides 30 minutes of operation when the external coil is disconnected. However, during normal operation, the pump is powered by a portable battery pack or by house current.

The pump has a stroke volume of 62 ml and is capable of pumping 8.5 l/min, (l0mmHg filling pressure, 120mmHg outlet pressure). Extensive mock loop testing has demonstrated progressive improvement in system reliability.

Twenty-six chronic animals have had circulatory support with an average period of pumping of 62 days, the longest period of support being eight months. Experiments were terminated in 18 animals for pump related problems (electromechanical 8, moisture related 6, sac rupture 4) and in 7 for animal related problems (thromboembolic 3, bleeding 2 and infection 2). One animal is ongoing.

Studies to date are very encouraging and suggest that, with further refinement, a reliable assist pump can be developed that will have important clinical application.

*By Invitation

TUESDAY MORNING, April 28, 1992

9:00 a.m. SCIENTIFIC SESSION - Los Angeles Ballroom

13. Aortic Valve Replacement With Pulmonary Homografts: Early Experience

GINO GEROSA, M.D.*, DONALD N. ROSS, F.R.C.S.

PETER E. BRUECKE, M.D., ANTONI J. DZIATKOWIAK, M.D.*,

SOPHIA MOHAMMAD, MSc* and

DINO CASAROTTO, M.D.*

Verona, Italy; Linz, Austria; Krakow, Poland;

London, England

The increased use of aortic homografts (AH) as aortic valve substitute and the limited availability of donor valves prompted us to consider the pulmonary homograft (PH) as an alternative substitute for aortic valve replacement (AYR). The aim of our study is to compare the morphologic, ultrastructural and biomechanical properties of PH leaflets with the AH leaflets and to present the early phase results using PH for AYR. Light, scanning and transmission electron microscopy have shown that PH leaflets are thinner than the aortic with a lesser content of elastic tissue in the ven-tricularis layer. Moreover there were not marked differences in the ultrastruc-ture. Uniaxiale tensile tests were carried out on 36 cusps from human pulmonary and aortic valves using an Instron test machine. The strain at 200 KPa was found to be similar for both pulmonary and aortic leaflets (approx 26%) cut radially. Circumferential strips appear to be more extensible in pulmonary leaflets, than in aortic (15% and 9% respectively). The ultimate tensile strength (UTS) for circumferential strips was found to be one and a half times as large for aortic when compared with pulmonary, but there was relatively little difference between the radial strips. As far as clinical experience is concerned, from September 1988 through September 1991, 107 consecutive patients 20-78 years old received either fresh-antibiotic or cryopreserved PH for AYR. The PH's were inserted in place of patients' diseased aortic valve with two different techniques: freehand in subcoronary position or as a "short cylinder" inside the aortic root. There,were 3 hospital deaths (2.8%). Follow-up was complete (1-36 mo.), all surviving patients have been followed with serial color flow Doppler echo.cardiography. There were no late deaths. 3 patients (2.9%) underwent reoperation because of severe aortic regurgitation (1,4 and 15 mo. post-op.) due to technical problem (mismatch in size between PH and aortic annulus) in 2 cases and probably due to graft rejection in 1 case. Mild aortic regurgitation has been detected in 3 patients (2.9%). No patients incurred in thromboembolicepisodes or infective endocarditis. According to our results the PH has shown to have ultrastructural and biomechanical properties similar to that of the AH. Furthermore PH has shown to be more pliable and easier to insert giving promising short term results.

*By Invitation

14. Infant Repair of Complete Atrioventricular Canal Defects: 20 Year Trends

FRANK L. HANLEY, M.D.*, RICHARD A. JONAS, M.D.,

JOHN E. MAYER, JR., M.D. and

ALDO R. CASTANEDA, M.D.

Boston, Massachusetts

Three hundred nineteen infants with complete atrioventricular canal defects have undergone surgical repair at our institution from 1972 to 1991. Analysis of these cases by 5 year intervals reveals a number of important institutional trends.

These trends show increasing volume and complexity of cases over time. In spite of this the number of cases undergoing primary repair increased and the need for early reoperation and the in hospital mortality have decreased markedly. The most recent 5 year experience reveals a 1.5% mortality in patients aged 0 to 6 months at the time of surgery and a 4.5% mortality in patients aged 7 to 12 months (overall 2.9% in hospital mortality).

These data support primary repair in infancy as the procedure of choice and suggest that repair before 6 months may provide further benefit. Further analysis suggests that precise surgical technique, primarily with avoidance of residual mitral regurgitation and to a lesser extent with avoidance of residual ventricular level shunting, is an important factor in the improved outcome and can be reliably achieved in early infancy.

*By Invitation

15. Deep Hypothermia and Circulatory Arrest: Determinants of Stroke and Early Mortality in 656 Adult Patients

LARS G. SVENSSON, M.D.*, E. STANLEY CRAWFORD, M.D.,

KENNETH R. HESS, M.S.*, JOSEPH S. COSELLI, M.D.*

and HAZIM J. SAFI, M.D.*

Houston, Texas

Deep hypothermia with circulatory arrest are being used more often for complicated cardiovascular surgery, particularly for repair of the aortic arch and acute aortic dissection. There are, however, few large studies that have documented the safety of this technique in adults. We have therefore evaluated our results for this type of adjunct to determine the independent predictors of early death and postoperative stroke by logistic regression analysis. Of the 656 patients operated upon between 7/7/79 and 1/30/91, 43% (N = 283) were female, median age was 64 years (range 10 to 88 years), 12% (N = 77) had acute dissection, 26% (N = 173) had previously had cardiac or ascending aortic surgery, and 13% (N = 85) had a history of cerebrovascular disease. The median circulatory arrest time was 31 minutes (range 7 to 120 minutes). The univariate predictors of transient or permanent stroke, either global or hemiparetic, which occurred in 44 patients (7%), were (p<0.05): increasing age, history of cerebrovascular disease, circulatory arrest time (7-29 minutes = 12/298 [4%]; 30-44 minutes = 15/201 [7.5%]; 45-59 minutes 9/84 [10.7%]; 60-120 minutes 7/48 [14.6%]), car-diopulmonary bypass time, and concurrent descending thoracic aorta repair. The independent determinants for stroke were (p < 0.05): history of cerebrovascular disease, previous aortic surgery distal to the left subclavian artery, and cardiopulmonary bypass time. A history of aortic valve incompetence was associated with a lower risk (adjusted odds ratio 0.42, p=0.015). The independent determinants for increased risk of early death, which occurred in 66 (10%) of patients, were (p < 0.05): increasing age, Mar-fan syndrome, concurrent distal aortic aneurysm, previous ascending aortic surgery, cardiopulmonary bypass time, cardiac complications, renal complications, and stroke. We conclude that deep hypothermia with circulatory arrest is a safe technique for the repair of complex aortic problems provided circulatory arrest time and cardiopulmonary bypass time is not excessive. Furthermore, the determinants of stroke and death are predominantly related to the patients clinical characteristics.

10:00 a.m. INTERMISSION - VISIT EXHIBITS

*By Invitation

10:45 a.m. SCIENTIFIC SESSIONS - Los Angeles Ballroom

16. Intra-Aortic Balloon Pumping in Cardiac Surgical Patients: Risk Analysis and Long Term Follow-Up

KEITH S. NAUNHEIM, M.D.*, MARC T. SWARTZ, B.A.*,

D. GLENNPENN1NGTON, M.D.,

GEORGE C. KAISER, M.D., LAWRENCE R. McBRIDE, M.D.*

and ANDREW C. FIORE, M.D.*

St. Louis, Missouri

The intraaortic balloon pump (IAB) is usually the first mechanical device inserted for perioperative cardiac failure, however, little current data is available regarding short and long-term effectiveness. From Jan. 1983 through Nov. 1990, 6,856 adult patients (pts) underwent cardiac surgical procedures, 580 of whom (8.5%) had lABs inserted preoperatively (preop) 107 pts, intraoperatively 419 pts, or postoperatively 54 pts. There were 374 males, 206 females with a mean age of 63.9 yrs (range 19-88). Operations included 336 CABG, 75 mitral, 54 aortic, 15 double valve replacements and 100 other procedures. Operative mortality for IAB pts was 44%. Univariate and multivariate analysis of 26 parameters revealed 6 independent predictors of mortality.

There were 96 (16.5%) lAB-related complications of which 43 required IAB removal for ischemia and 53 required surgical intervention. lAB-related surgical procedures included thrombectomy (34), vascular repair (15), aortic repair (2) and fasciotomy (2). Univariate analysis demonstrated no relationship between any of the lAB-related complications and survival. Only 34 of the 580 pts (6%) were lost to follow-up. There were 75 late deaths, the etiology of which was cardiac in 41 (55%), noncardiac in 20 (27%) and unknown in 14 (19%). Actuarial survival at 1, 5 and 9 years is 51%, 42% and 33%. Of the 217 hospital survivors still alive and contacted, 81% were in NYHA Class I (114) or II (60).

These data show 1) operative mortality for pts requiring IAB in the perioperative period remains high; 2) perioperative risk factors can be identified; 3) IAB complications do not effect survival; 4) operative survivors can achieve prolonged survival with excellent functional results; and 5) consideration for alternative methods of circulatory support is justified.

*By Invitation

17. The Effect of Coronary Reoperation on the Survival of Patients With Stenoses in Saphenous Vein to Coronary Bypass Grafts

BRUCE W. LYTLE, M.D., FLOYD D. LOOP, M.D.,

PAUL C. TAYLOR, M.D.*, ROBERTONOVOA, M.D.*,

MARLENE GOORMASTIC, M.P.H.* and

DELOS M, COSGROVE, M.D.

Cleveland, Ohio

Coronary reoperations (reop) have not yet been shown to improve survival. To examine the question of whether coronary reop improves the survival of patients with stenoses in saphenous vein to coronary bypass grafts (SVG), we retrospectively reviewed 1117 patients who had coronary bypass surgery then underwent a postoperative coronary angiogram (stenotic cath) that documented a stenosis (20-99%) of at least one SVG. Reop within one month of the stenotic cath was performed for 394 patients (reop group) whereas 723 patients (med group) received initial medical treatment (no reop or PTCA within 1 year). Compared with the med group, patients in the reop group were older, more symptomatic, had a higher incidence of left main stenosis and fewer patent bypass grafts (all p < 0.001). In-hospital mortality for the reop group was 3.8%. Mean post cath follow-up of the entire group was 80 months.

Based on the interval between the primary operation and the stenotic cath, patients were designated as having early (< 5 yrs.) or late (> 5 yrs.) SVG stenoses. Univariate and multivariate analyses were used to identify factors influencing the survival of these subgroups. Reoperation was not identified as a variable improving the survival of patients with early SVG stenoses.

For patients with late SVG stenoses, moderate or severe impairment of left ventricular function, advanced age, triple-vessel or left main stenosis and stenosis in a SVG to the left anterior descending artery (LAD) (all p < 0.001) decreased survival while reoperation (p = 0.0015) improved survival. The benefit of reop was very strong for patients with SVG-LAD stenoses with survival of 84% and 74% for the reop group versus 76% and 53% for the med group at 2 and 4 post cath years, respectively (p = 0.004). Even for patients with Class I or II symptoms, reoperation prolonged survival (p = 0.02 with multivariate testing). This is the first study showing that coronary reoperation improves the survival of any patient subset and documents enhanced survival with reoperation for patients with late SVG stenoses, particularly those with SVG-LAD stenoses.

11:25 p.m. ADDRESS BY HONORED SPEAKER

Coronary Artery Bypass Graft Surgery; Twenty-five Years Later. Some Landmarks

Rene G. Favaloro, M.D., Buenos Aires, Argentina

12:10 p.m. ADJOURN FOR LUNCH - VISIT EXHIBITS

CARDIOTHORACIC RESIDENTS' LUNCHEON - Century Room

*By Invitation