TUESDAY MORNING, MAY 5, 1998

7:00 a.m. C. WALTON LILLEHEI RESIDENT

FORUM SESSION

supported by an educational grant from

st. jude medical, inc.

Ballroom, Hynes Convention Center

Moderators: Eric A. Rose, M.D.

Andrew S. Wechsler, M.D.

F1. MYOCARDIAL TISSUE ENGINEERING WITH AUTOLOGOUS MYOBLAST IMPLANTATION.

Julia Dorfman, M.D.*, Minh Duong, B.Sc.*, Audrius Zibaitis, M.D.*, M.Sc., Marc Pelletier, M.D.* and Ray C.-J. Chiu, M.D., Ph.D.

Montreal, Quebec, Canada

Implanting autologuos skeletal muscle-derived myoblasts, i.e. satellite cells, for myocardial replacement has many advantage, compared to other approaches of implanting either fetal cardiac myocytes (ethical and donor availability issues) or established cell lines (oncogenicity). Furthermore, autologous myoblasts do not require immunosuppression. The feasibility of satellite cells implanted into the myocardium differentiating into muscle fibers was confirmed in this study using a unique cell labeling technique.

Methods: Myoblasts are isolated from the skeletal muscle of adult rats, and cultured in vitro to proliferate the number of cells. These cells are labelled with DAPI (4', 6-diamidino-2-phenylindole) which binds to the cellular DNA and to the protein tubulin, and form a fluorescent complex. One million labeled satellite cells are implanted into the myocardium of rats isogenic to the donors. The cells were injected into the left ventricular wall with the heart exposed. The cardiac specimens are harvested 1 to 4 weeks following myoblast implantation. Histological sections are examined under fluorescent microscopy.

Results: By examining the cell culture plates under fluorescent microscope, the labeling efficiency of satellite cells with DAPI is found to be nearly 100%. In 8 rat hearts, the progressive differentiation of implanted myoblasts from isolated cells to fully developed striated muscle fibers can be observed. These new muscle fibers clearly labeled with DAPI fluorescence grew adjacent to unlabelled native cardiac muscle fibers.

Discussion: Our earlier studies of implanting autologous canine satellite cells into cryo-injured myocardium indicated that these cells could differentiate into cardiad phenotype myocytes. The presence of intercalated disks in these fibers suggested that they could fuse with native cardiac fibers to supplement ventricular function. However, it had been difficult to firmly establish these findings using cell markers, proving these neo-myocardium had indeed been derived from the implanted myoblasts. This study shows that DAPI has excellent labeling efficiency for satellite cells in vitro and in vivo. For the first time, we are able to confirm that the satellite cells implanted into the myocardium did in fact differentiate into fully developed, labeled muscle fibers.

Conclusions: Because of the obvious advantages of using autologous donor myoblast cells, clinical application of this approach appears desirable, and may provide a novel strategy for the management of heart failure patients in the future.

*By invitation

F2. COMPLETE REVERSAL OF ISCHEMIC WALL MOTION ABNORMALITIES BY COMBINED USE OF GENE THERAPY WITH TRANSMYOCARDIAL LASER REVASCULARIZATION.

Umer Sayeed-Shah, M.D.*, Michael J. Mann, M.D.*, Jeffrey Martin, M.D.*, Sergey Grachev, M.D.*, Sharon Reimold, M.D.*, Rita Laurence, B.S.*, Victor Dzau, M.D.* and Lawrence H. Conn, M.D.

Boston, Massachusetts

Introduction: Transmyocardial Laser Revascularization (TMR) results in symptomatic improvement in patients with chronic ischemic heart disease. This effect is thought to result from an angiogenic response to TMR in the ischemic myocardium. Preliminary data from our laboratory indicate that TMR enhances both the efficiency of gene delivery and the degree of transgene expression with the direct myocardial injection of non-viral vectors. We hypothesized that direct myocardial injection of plasmid DNA encoding the gene for vascular endothelial growth factor-1 (VEGF-1) could enhance the revascularization achieved by TMR in ischemic hearts, resulting in improved cardiac function.

Methods: 29 Yorkshire pigs underwent the placement of an ameroid constrictor at the proximal left circumflex artery via a left mini-thoracotomy, and were allowed to recover. Group I, (Ischemic controls, n = 5) had no further intervention. Group II, (n = 4) underwent TMR in the area at risk at 6 weeks. Group III, (n = 5) underwent TMR at 6 weeks with 3 equidistant intramyocardial injections of 100µg of an expression plasmid encoding the gene for β-galactosidase (pSV-β-gal) surrounding each TMR site. Group IV, (n = 4) had sets of 3 equidistant injections of l00 µg of an expression plasmid encoding the gene for VEGF-1, (pSV-VEGF) without TMR at 6 weeks. Group V, (n = 5) had TMR with each site surrounded by 3 injections of 100µg of pSV-VEGF. All animals were harvested at 12 weeks. Six additional pigs underwent TMR with either pSV-0-gal or pSV-VEGF injections, and were harvested at 8 weeks.

Results: Left ventricular free-wall motion by transesophageal and epicardial echocardiography was assessed by a cardiologist in a blinded manner. Hearts were scored as normal (no regional wall motion abnormality, normal systolic function), or abnormal. All (0/5) of the ameroid alone hearts were abnormal, whereas 75% (3/4) of the TMR hearts, 60% (3/5) of the TMR-β-gal hearts, and 50% (2/4) of the VEGF hearts displayed evidence of persistent wall motion abnormality. In contrast, all (5/5) of the heart treated with TMP-VEGF displayed no evidence of wall motion abnormality. Only the TMR-VEGF hearts had a statistically significantly different rate of wall motion abnormality compared to untreated ischemic hearts (p = 0.004 by two-tailed Fisher's Exact Test). Of note, 2 out of 3 TMR-VEGF pigs examined at 2 weeks post-treatment had completely normal ventricular function, whereas 3 TMR-p-gal pigs examined at the same time point all had persistent abnormalities (p>0.05).

Conclusion: These results suggest that the combined use of TMR with direct injection of an expression plasmid encoding VEGF-1 completely reverses ischemic wall motion abnormalities at 6 weeks after therapy, and that the resolution of wall motion abnormalities may occur as early as 2 weeks after therapy.

*By invitation

F3. TRANSMYOCARDIAL LASER REVASCULARIZATION FAILS TO PREVENT LEFT VENTRICULAR FUNCTIONAL DETERIORATION AND ANEURYSM FORMATION AFTER ACUTE MYOCARDIAL INFARCTION IN SHEEP.

Ramin Malekan, M.D.*, Scott T. Kelley, M.D.*, Yasuyuki Suzuki, M.D.*, Carol Reynolds, M.D.*, Theodore Plappert, C.V.T.*, Martin St. John-Sutton, M.D.*, L. Henry Edmunds, Jr., M.D. and Charles R. Bridges, M.D., D.Sc.

Philadelphia, Pennsylvania

Transmyocardial laser revascularization (TMLR) seeks to nourish left ventricular (LV) myocardium from cavity blood. We tested the hypothesis that TMLR prevents subsequent LV functional deterioration and aneurysm formation after acute antero-apical myocardial infarction.

In 21 anesthetized Dorset-hybrid sheep, snares were placed around the distal left ant. descending and second diagonal coronary arteries and a flow probe was fitted around the ascending aorta. After baseline hemodynamic measurements and transdiaphragmatic echocardiograms 10 days later, myocardium supplied by snared arteries was perforated using a CO² laser (4 channels/cm²; 29-35 transmural holes) immediately prior to infarction in 11 sheep. Ten control animals did not have TMLR. Tightened snares infarcted 23% of the LV mass. Hemodynamic and echocardiographic studies were repeated one hour and 2, 5 and 8 weeks after infarction.

Control animals developed large antero-apical aneurysms confirmed at autopsy. At 8 weeks stroke work decreased from 253 ± 78 to 139 ± 54 ergs x 10³ (p<0.01); cardiac output decreased from 2.4 ± 0.7 to 1.6 ± 0.7 1/min (p<0.01); ejection fraction decreased from 47.4 ± 8.6 to 23.4 ± 5.8 % (p<0.01); left ventricular end diastolic pressure increased from 1.7 ± 1.0 to 5.7 ± 1.0 mmHg (p<0.05); end systolic volume increased from 28.3 ± 6.9 to 75.9 ± 18.8 ml (p<0.01) and end diastolic volume increased from 54.0 ± 11.3 to 98.3 ± 16.0 ml (p<0.01). Mean arterial and central venous pressures did not change significantly. Histology two days after TMLR showed holes filled with fibrin and scattered red cells. At 5 and 8 weeks the infarct consisted of dense collagen, fibroblasts and lymphocytes with occasional islands of viable myocytes in both groups. Laser holes could not be found. No significant (p<0.05) differences were found between TMLR and control sheep for any measurement at any time point after infarction; histologic sections also did not differ between groups at 5 and 8 weeks. We conclude that TMLR does not revascularize acute myocardial infarction sufficiently to attenuate LV functional deterioration and aneurysm formation.

*By invitation

F4. METHYLPREDNISOLONE REDUCES THE INFLAMMATORY RESPONSE TO CARDIOPULMONARY BYPASS IN NEONATES: TIMING OF DOSE IS IMPORTANT.

Andrew J. Lodge, M.D.*, Paul J. Chai, M.D.*, C. William Daggett, M.D.*, Ross M. Ungerleider, M.D. and James Jaggers, M.D.*

Durham, North Carolina

Introduction: Outcome after neonatal heart surgery is sometimes hampered by a severe inflammatory response to cardiopulmonary bypass (CPB). This can present as total body edema or pulmonary dysfunction. This study was designed to evaluate the efficacy of preoperative methylprednisolone (MP) administration in limiting this response and compares the effect of giving MP eight hours preoperatively to the more clinically common practice of adding MP to the CPB circuit prime. Methods: Three groups of neonatal piglets were studied. A control group (Control, n = 6) received no preoperative medication. One experimental group (PreOp-MP, n = 6) received methylprednisolone sodium succinate (30 mg/kg) both 8 hours and immediately preoperatively. Another experimental group (Prime-MP) received no preoperative treatment, but MP (30 mg/kg) was added to the CPB circuit prime. All animals underwent CPB and 45 minutes of deep hypothermic circulatory arrest. Hemodynamic and pulmonary function data were acquired prior to CPB (BSL) and at 30 (Post30) and 60 (Post60) minutes after CPB. Results: Post CPB lung water content was significantly lower in the PreOp-MP group compared to the control group (p = 0.001) and the Prime-MP group (p = 0.026). Total body water gain was significantly less in the PreOp-MP group compared to the control group (p = 0.003) and the Prime-MP group (p = 0.01). Pulmonary function data are presented in the following table (results for each parameter were not significantly different at Post60 compared to Post30):

	Stage	Control	Prime-MP	PreOp-MP
Compliance	BSL	3. 05 ± 0.24	2.99 ± 0.22	3. 28 ± 0.28
(ml/cm H2O)	Post30	1.59 ± 0.26*	2.43 ± 0.28*†	3.14 ± 0.30†‡
A-a Gradient	BSL	132 ± 12	139 ± 22	112 ± 10
(mmHg)	Post30	396 ± 49*	241 ± 31†	179 ± 25†
PVR	BSL	2503 ± 674	3268 ± 345	1766 ± 377
(dyne-sec-cm-5)	Post3 0	11421 ± 787*	7715 ± 1690*	3775 ± 429†‡
PVR: pulmonary vascular resistance; All results expressed as mean ± standard deviation; p values by ANOVA
*:p<0.01 vs. BSL †:p < 0.01 vs. Control ‡:p < 0.05 vs. Prime-MP

Conclusions: MP given eight hours and immediately before surgery provides superior protection from the inflammatory response to CPB compared to no treatment and to addition of MP to the CPB circuit prime. These results suggest a simple, available, cost effective means of reducing early post-CPB morbidity, especially if used in selected high risk patients.

*By invitation

F5. CONTINUOUS PERFUSION OF DONOR HEARTS IN THE BEATING STATE EXTENDS PRESERVATION TIME AND IMPROVES RECOVERY OF FUNCTION.

Waleed H. Hassanein, M.D.*, Lambros Zellos, M.D.*, Tracey A. Tyrrell, B.A.*, Nancy A. Healey, B.S.*, Michael D. Crittenden, M.D.*, Vladimir Birjiniuk, M.D.* and Shukri F. Khuri, M.D.

West Roxbury, Massachusetts

Improving methods of donor heart preservation may permit prolonged storage and remote procurement of cardiac allografts. We hypothesized that continuous sanguineous perfusion of the donor heart in the beating working state may prolong myocardial preservation. We developed a simple, portable perfusion apparatus that may be used for donor heart preservation. Contractile, metabolic, and vasomotor functions were monitored simultaneously in an isolated pig heart. Metabolic function was monitored by myocardial tissue pH. Vasomotor function was assessed in isolated coronary ring chambers. Hearts were randomized into three Groups: I (n = 5) cardioplegic arrest, 12hr storage @ 4°C using Modified Belzer's Solution, and 2hr sanguineous reperfusion in the working state, or II (n = 6) 12hr continuous perfusion in the beating working state, 30min arrest (to simulate time needed for re-implantation), and 2hr reperfusion, as above. Group III (n = 7) served as coronary rings controls. Results (m ± SD): At 2 hours of reperfusion LV developed pressure in Group II was higher than in I (90+6, 53+15mmHg, P = 0.005). Significantly less myocardial edema was observed in Group II vs I (73 ± 4, 80 ± 1% H₂0 content, P = 0.01). Significantly less myocardial acidosis was noted in Group II vs I during preservation (pH 7.3 ± 0.01, 6.1 ± 0.03, P<0.001) and reperfusion (pH 7.3 ± 0.008, 6.8 ± 0.05, P<0.001).Coronary endothelial vasomotor function was better preserved in Group II vs I as evidenced by the dose response relaxation of coronary rings to 10-⁸M Bradykinin (37%, 55%∆ baseline, P = 0.01). Conclusion: This new method of heart preservation extends the current preservation limit and avoids time dependent ischemic injury, thereby allowing for distant procurement of donor organs.

*By invitation

F6. TRANSGENE EXPRESSION FOLLOWINGADENOVIRAL-MEDIATED RETRANSFECTION OF RAT LUNGS IS INCREASED AND PROLONGED BY TRANSPLANTATION-LEVEL IMMUNOSUPPRESSION.

Stephen D. Cassivi, M.D.*, Mingyao Liu, M.D.*, Annette Boehler, M.D.*, Keith Tanswell, M.B.*, Arthur Slutsky, B.A.Sc., M.A.Sc., M.D.* and Shaf H. Keshavjee, M.Sc., M.D.*

Toronto, Ontario, Canada

Sponsored by: Thomas R.J. Todd, M.D., Toronto, Ontario, Canada

The potential benefits of gene therapy, in modifying donor organs to better withstand the process of transplantation, have yet to be fully realized. In vivo gene transfer using adenoviral vectors has had limited success due to a host immune response inducing severe inflammation, which limits both the amount and duration of transgene expression, and obviates effective retransfection. Since this transgene expression diminishes over time, the ability to achieve effective retransfection is essential to providing successful chronic gene therapy. Our recent studies have shown that by administering transplant-level immunosuppression, transgene expression, following initial adenoviral-mediated transfection, can be significantly increased and prolonged for up to 5 weeks. In our current studies, we have shown that by administering transplant-level immunosuppression consisting of cyclosporine, azathioprine, and methylprednisolone, effective and prolonged adenoviral-mediated retransfection of rat lungs can also be achieved. Male Lewis rats, randomly assigned to immunosuppressed or non-immunosuppressed (control) groups, received intratracheal delivery of a first-generation, replication-deficient adenovirus containing the reporter gene β-galactosidase at day 0 and again at day 35. Whereas control rats had virtually no detectable transgene expression following retransfection (2.8 ± l.lRLU/ng protein, 0.4 ± 0.1 RLU/ng protein, and 1.2 ±0.2 RLU/ng protein, at days 1, 7 and 14 respectively), immunosuppressed rats demonstrated significantly elevated transgene expression, peaking at 7 days and showing prolonged and elevated expression at 2 weeks (188.4 ±46.4 RLU/ng protein, 294.2 ±43.7 RLU/ng protein, and 183.8 ± 53.6 RLU/ng protein, at days 1, 7 and 14 respectively; p < 0.005 at all three time points). To confirm these quantitative chemiluminescence results, we performed chromogenic staining of lung tissue using the X-gal stain for the marker gene β-galactosidase, which showed a diffuse pattern of transgene expression in the lung tissue of immunosuppressed rats. Conversely, virtually no positive staining was observed in the control group. Chronic gene therapy, through effective gene retransfection, previously limited by a severe immune response, can therefore be achieved with transplant-level immunosuppression. Improved outcomes in lung transplantation may now be possible, through enhanced and prolonged genetic modification and gene therapy retreatment.

*By invitation

F7. ALTERED MYOCARDIAL GENE EXPRESSION FOLLOWING BRAIN DEATH.

Thomas Yeh, Jr., M.D, Ph.D.*, Andrew S. Wechsler, M.D., Laura J. Graham, L.V.T.*, Kathryn E. Loesser, Ph.D.*, Domenic A. Sica, M.D.*, Luke Wolfe, Ph.D.* and Emma R. Jakoi, Ph.D.*

Toronto, Ontario, Canada and Richmond, Virginia

Objectives: The depressed myocardial function observed in brain dead organ donors has been attributed to massive sympathetic discharge and catecholamine cardiotoxicity. Although physiologic and morphologic changes associated with brain death have been reported, the molecular mechanisms underlying the observed hemodynamic instability are poorly understood. Because chronic catecholamine administration alters gene expression in myocardial cell culture, and because humans with elevated systemic catecholamines manifest decreased expression of myocardial genes important in contractility, we tested the hypothesis that activation of the neuromyocardial axis modulates the expression of genes important for contractility.

Methods: In a balloon expansion model of increased intracranial pressure (↑ICP) in rabbits (n = 23), sympathetic regulation of LV myocardial gene expression was analyzed. At timed intervals after brain death, mean arterial pressure, heart rate, ECGs and histologic myocardial injury were compared with sham-operated controls. Systemic catecholamine levels were correlated with changes in hemodynamic parameters in response to ↑ICP. Using RNA blot hybridization analysis, the effect of ↑ICP on levels of mRNA encoding myofilaments, adrenergic receptors, sarcoplasmic reticulum proteins, and stress induced genes [i.e. immediate early genes and heat shock protein 70 (hsp70)] was measured.

Results: In contrast with sham operated controls, ↑ICP resulted in several statistically significant changes, including an immediate pressor response that correlated temporally with diffuse ECG ST segment changes and histologic injury. A concurrent 5-8 fold increase in systemic epinephrine and norepinephrine levels dropped precipitously below baseline to near zero levels within 2 hours. Four hours after ↑ICP, levels of mRNA encoding skeletal and cardiac α-actins, egr-1, and hsp70 were selectively increased over sham operated controls.

Conclusions: This study of ↑ICP is the first systematic evaluation of multiple programs of myocardial gene expression in an in vivo model of brain injury, mRNAs encoding egr-1, hsp70 and α-actins are all significantly increased after ↑ICP, a pattern which differs from those typically associated with depressed myocardial contractility and chronic elevation of systemic catecholamines. Myocytes in experimental brain deal undergo a molecular transformation that may have important implications when such hearts are donated for transplantation.

*By invitation

F8. FACTOR IXai IS A SELECTIVE ANTICOAGULANT AGENT FOR CARDIOPULMONARY BYPASS WHICH PROVIDES EXTRACORPOREAL CIRCUIT ANTICOAGULATION WITH PRESERVATION OF SURGICAL HEMOSTASIS.

Talia B. Spanier, M.D.*, †Mehmet C. Oz, M.D., David M. Stern, M.D.*, Niloo M. Edwards, M.D.*, Eric A. Rose, M.D. and Ann Marie Schmidt, M.D.*

New York, New York

INTRODUCTION: The mechanism of activation of coagulation in the setting of cardiopulmonary bypass (CPB) remains to be delineated. We believe that activation of coagulation in the intravascular space/extracorporeal circuit is due to activation of the intrinsic pathway, while activation in the extravascular space / surgical wound is separately due to the extrinsic / tissue factor (TF) mediated pathway. An ideal anticoagulant agent in CPB, therefore, would be one which would SELECTIVELY inhibit CPB circuit / contact-mediated thrombosis, while preserving extravascular / TF-mediated hemostasis. We focused on Factor IX due to its unique location in the intrinsic pathway of the coagulation cascade and hypothesized that blockade of Factor IX would achieve this end.

METHODS: Active-site blocked Factor IXa (IXai), a competitive inhibitor of the assembly of IXa in the X activation complex, was prepared by enzymatic modification of IXa with dansyl-glutamyl, glycyl arginyl chlormethylyketone. CPB was performed in 12 baboons (7 with IXai (460µg/kg)/no reversal, 5 with heparin/protamine (H)). After 1 hour at 32°C, baboons were weaned from CPB and observed for 3 hours postoperatively. Intraoperative blood and tissue samples were taken for analysis of thrombin generation, macrophage procoagulant activity and immunohistochemistry.

RESULTS: Normal circuit pressures were maintained throughout CPB in both groups with no clot formation in the tubing. Scanning EM of arterial filters confirmed no differences in fibrin / platelet deposition. Blood loss was significantly less with IXai vs H (480 ± 75 vs 1150 ± 115ml, p<0.05). A modified bleeding time, used to assess extravascular hemostasis, revealed that there was no evidence of enhanced extravascular bleeding during CPB or postoperatively compared to baseline in IXai treated animals. In contrast, significantly increased extravascular bleeding was observed after infusion of H during CPB, which returned to baseline after H reversal with protamine (p<0.05). In order to delineate the sites and extent of thrombin generation in CPB, examination of plasma prepared from peripheral blood from animals treated with IXai compared to H revealed significantly increased levels of thrombin antithrombin III complex [TAT] and prothrombin activation complex (F1+2), with peak levels occurring on CPB, and then declining immediately after termination. Furthermore, retrieval of monocytes from baboon peripheral blood, arterial filter and CPB pump (representing the intravascular/ extracorporeal space) during the procedure revealed no demonstrable TF activity in IXai or H treated animals. In contrast, monocytes isolated from pericardial blood (representing the extravascular space) in IXai treated animals manifested significantly enhanced procoagulant activity. Immunohistochemical examination of the pericardium during CPB confirmed enhanced TF activity in IXai but not H treated animals with striking TF expression in the pericardial mesothelial cells, in macrophages, as well as in the vasculature and macrophages that pervade pericardial tissue.

CONCLUSIONS: The finding that overall thrombin generation is greater in baboons treated with IXai supports our hypothesis that separate mechanisms are responsible for activation of coagulation in CPB since, while effective antithrombotic effect is realized in the extracorporeal circuit with IXai, extravascular hemostasis is preserved. Activation of coagulation in the setting of CPB, therefore, is separately attributable to the intrinsic pathway in the bypass circuit/ intravascular space, while the extrinsic /TF pathway is predominantly responsible for that within the surgical wound. Selective anticoagulation with IXai is a novel anticoagulant for CPB which affords a mechanism by which to study activation of coagulation in CPB, and also selectively prevents intravascular/extracorporeal circuit thrombosis with preservation of surgical hemostasis.

†1994-96 AATS Robert E. Gross Research Scholar

*By invitation

9:00 a.m. PLENARY SCIENTIFIC SESSION

Ballroom, Hynes Convention Center

Moderators: Floyd D. Loop, M.D.

James L. Cox, M.D.

14. EXCELLENT LONG TERM FUNCTIONAL OUTCOME AFTER SURGERY FOR ANOMALOUS LEFT CORONARY ARTERY FROM THE PULMONARY ARTERY.

Andrew D. Cochrane, F.R.C.S.*, David M. Colmenan, M.B.Ch.B.*, Christian P.R. Brizard, M.D.* and Tom R. Karl, M.D.

Melbourne, Australia

Discussant: Yasuharu Imai, M.D., Shinjuku-ku, Japan

During 1980-1996, 20 patients underwent surgery for this condition. In 19 patients the left main coronary artery arose from the MPA or RPA, and in one infant the circumflex alone came from the MPA. The median age at presentation was 9 months (range 6 weeks to 26 years). Fourteen patients were severely symptomatic with heart failure at time of operation, at a younger median age of 6 months. Mitral regurgitation was present in all except two patients, usually moderate to severe in degree; one other patient had undergone a mitral valve replacement before the diagnosis of ALCAPA was made. One infant had LV anterior wall rupture and a hemopericardium due to transmural infarction. Treatment involved Takeuchi operation (n=12), direct reimplantation (n = 6), aorto-coronary bypass in an older patient after previous ligation (n = 1) and ligation of the isolated anomalous circumflex artery (n = 1). There were no operative deaths (0%, CL = 0-17%) and no late deaths. Although 5 patients required LVAD support, their outcome has been excellent with no increment in early mortality or late sequelae (p>0.05). There have been two late reoperations, one for supravalvar PA stenosis after Takeuchi repair, and one mitral valve replacement. Nine patients followed up at this hospital have undergone late studies, including three who required LVAD and one who had a late mitral valve replacement. On nuclear gated scan at a mean of 6 years from surgery the LVEF was 64 ± 3 % at rest and increased normally to 74 ± 3 % on exercise. First pass RVEF was normal at 65 ± 3 %. On echocardiography 6 patients have mild or no mitral regurgitation, while 3 patients have persistent moderate or severe regurgitation with one infant requiring medical therapy. The LV dimensions, fractional shortening and indices of systolic and diastolic function are no different between the patients and healthy age-matched controls (p>0.05). In the 7 children old enough to perform a treadmill exercise test, the exercise time and blood pressure response were normal. Long-term outlook and late LV function is good, even in infants with severely impaired LV function at presentation or in those who require LVAD support. Cardiac transplantation should rarely be necessary in this group, even in patients with initial severe impairment of LV function.

*By invitation

15. EFFECT OF DONOR AGE AND DONOR ISCHEMIC TIME ON INTERMEDIATE-TERM SURVIVAL AND SECONDARY ENDPOINTS AFTER LUNG TRANSPLANTATION.

Dan M. Meyer, M.D.*, Leah H. Bennett, Ph.D.*, Richard J. Novick, M.D. and Jeffrey D. Hosenpud, M.D.*

Dallas, Texas; Richmond, Virginia; London, Ontario, Canada and Milwaukee, Wisconsin

Discussant: G. Alexander Patterson, M.D., St. Louis, Missouri

Objective: Pressure to expand the donor pool has required the use of lungs from older donors or from more distant procurement areas. The long-term consequences of this policy have not yet been fully addressed. The effect of donor age and donor ischemic time on intermediate survival and important secondary endpoints after lung transplantation was therefore determined. Methods: A cohort of 1450 lung transplant recipients with complete 2 year followup, operated upon in the United States between April 1, 1993 and March 31, 1996, were studied. Results: As shown below, donor age when analyzed independently did not affect intermediate survival or the secondary endpoints (p>0.05 for all variables).

Similarly, when the interaction between ischemic time and donor age was examined in all of the multivariate models, none of the secondary endpoints was found to be significantly influenced. However, the combined interaction between donor age and ischemia time demonstrated a significantly worse survival at two years (p = 0.02) at donor age of > 50 years and donor ischemic time > 7 hours. Conclusions: Donor age and donor ischemic time did not negatively influence survival or important secondary endpoints after lung transplantation. However, intermediate-term survival was affected by the use of older donors when combined with a prolonged ischemic time. The impact of this combination should be considered when attempting to expand the donor pool.

*By invitation

16. VOLUME-OUTCOMES STUDIES OF CARDIOVASCULAR PROCEDURES IN NEW YORK STATE (1990-1995).

Keith Reemtsma, M.D., Eileen P. Shields, B.A.*, Jonathan M. Chen, M.D.*, Patrick J. Roonan, M.S.* and Annetine C. Gelijns, Ph.D.*

New York and Albany, New York

Discussant: Jack M. Matloff, M.D., Los Angeles, California

Background: Although conventional wisdom holds that increased experience yields better outcomes, studies have shown varied patterns in the relationship between volume of surgical cases and outcome. We have examined the relationship of hospital volume on in-hospital mortality in three patient cohorts: (1) adults undergoing coronary artery bypass graft (CABG) surgery without additional procedures (a high volume, State regulated procedure) (2) elective non-ruptured abdominal aortic aneurysm (AAA) repair (a low volume, unregulated procedure), and (3) pediatric patients undergoing cardiac surgery involving cardiopulmonary bypass (a low volume, regulated procedure). Methods: All cardiovascular procedures performed in New York State from 1990-1995 were evaluated in this study. Logistic regression χ²analysis (with resultant odds ratios) was performed evaluating hospital case volume with in-hospital mortality. Patients in the pediatric cohort were subdivided by age into three strata for analysis: <30 days (neonates), 31 days-1 year, 1-12 years of age, 13 -20 years of age.

Results: 109,372 patients underwent CABG surgery, 9,981 patients underwent AAA repair, and 6,618 pediatric patients underwent open-heart procedures. No significant association (p>0.05) was demonstrated between hospital case volume and mortality for CABG patients. However, a significant inverse relationship was noted (p = 0.0001) for patients undergoing elective AAA surgery, and children less than one year of age undergoing cardiac surgery. Additionally, for neonates, hospitals with case volumes below 100 patients displayed mortality rates three-fold higher than in the highest volume hospital (p = 0.0001). Patients 1-20 years of age did not show a significant inverse relationship by logistic regression analysis (p>0.05).

In New York State, cardiac surgery is regulated by a certificate-of-need process to approximately 30 hospitals state-wide, and the median CABG case volume for 1990-1995 was 2501 operations. By contrast, surgery for abdominal aortic aneurysms was performed in 195 hospitals, and in 138 hospitals fewer than 50 aneurysm operations were done over the six year period. Only three hospitals had operative volumes exceeding 300 operations in the same time period (1990-1995).

Conclusions: Although no relationship was demonstrable between hospital volume and mortality for patients undergoing CABG surgery, a strong correlation was noted for adults undergoing elective AAA repair and children undergoing open-heart procedures, most notably neonates. Neither AAA nor pediatric cardiac procedures were concentrated in hospitals with high volumes. These data support the concept of concentrating high risk procedures (especially those of low volume) in a limited number of hospitals that can provide an acceptable annual volume of activity.

10:00 a.m. INTERMISSION - VISIT EXHIBITS

*By invitation

10:45 a.m. PLENARY SCIENTIFIC SESSION

Ballroom, Hynes Convention Center

Moderators: Floyd D. Loop, M.D.

James L. Cox, M.D.

17. CRYOPRESERVED ARTERIAL HOMOGRAFTS IN THE TREATMENT OF MAJOR VASCULAR INFECTION: A COMPARISON WITH CONVENTIONAL SURGICAL TECHNIQUES.

Paul Robert Vogt, M.D.*, Hans-Peter Brunner-LaRocca, M.D.*, Thierry Carrel, M.D.*, Ludwig K. von Segesser, M.D.*, Christian Ruef, M.D.*, Wolfgang Kiowski, M.D.* and Marko Turina, M.D.

Zurich, Switzerland

Discussant: Tirone E. David, M.D., Toronto, Ontario, Canada

Objectives: The results with cryopreserved heart valve homografts in the treatment of infectious endocarditis suggest that the use of cryopreserved arterial homografts may improve the outcome in patients with major vascular infections.

Methods: Between 1990 and 1997, 72 patients with mycotic aneurysm (n = 29) or infected vascular prostheses (n = 43) of the thoracic (n = 26) or abdominal aorta (n = 46) were treated either with conventional surgery (n = 38) or implantation of a cryopreserved arterial homograft (n = 34). Survival, survival free of reoperation, lengths of intensive care, hospitalization, antibiotic treatment and costs were assessed. In patients with homografts, computed tomography, magnetic resonance imaging-angiography, or intravenous digital subtraction angiography were performed after a mean follow-up of 27 ± 16 months.

Results:	Homografts	Prosthesis (n=38)	p- value
	(n=34)
30-day-mortality	5.9%	18.4%	0.16
Total mortality	11.8%	31.8%	0.07
Infection eliminated	91%	53%	0.001
Reoperation during first postoperative year	9%	34%	0.01
Reoperation during follow-up time	9%	45%	0.001
Mean (± SD) time on respirator postoperative	3.5(6.1)	12.8(15.2)	0.001
(days)
Mean (± SD) time on intensive care (days)	5.6(9.7)	23.5 (29.5)	0.001
Mean (± SD) duration of hospitalization (days)	44.5 (28.6)	93.4(58.3)	0.001
Mean (± SD) duration of antibiotic therapy (days)	40.4(12.2)	159.7(138.3)	0.001
Cost of treatment per case	88,539 US$	310,979 US$	0.001

After 4 years, freedom from reoperation and death was 75 ± 10% for homografts and 34 ± 9% for prosthesis (p = 0.0018). Neither false aneurysm formation, stenosis, leakage, nor dilatation of homografts were observed.

Conclusions: The use of cryopreserved arterial homografts provides a safer, cheaper and more effective treatment for mycotic aneurysms and infected vascular prostheses than conventional surgical techniques.

*By invitation

18. NEUROPSYCHOLOGICAL OUTCOME FOLLOWING DEEP HYPOTHERMIC CIRCULATORY ARREST.

David L. Reich, M.D.*, M. Arisan Ergin, M.D., Suzan Uysai, Ph.D.*, Martin Sliwinski, Ph.D.*, JockN. McCullough, M.D.*, Jan D. Galla, M.D.*, Wayne Gordon, Ph.D.*, Mary Hibbard, Ph.D.* and Randall B. Griepp, M.D.*

New York, New York

Discussant: Richard A. Jonas, M.D., Boston, Massachusetts

Introduction: There is compelling evidence in the pediatric populations that prolonged periods of deep hypothermic circulatory arrest (DHCA) are associated with long term deficits in cognitive and some intellectual functions. Although DHCA is widely used in surgery of the thoracic aorta there is a paucity of such information in adult patients. Evaluation of neurological outcome traditionally has been limited to reports of the incidence of postoperative stroke. The incidence of and factors associated with long term neuropsychological dysfunction are unknown.

Methods: Under an IRB-approved protocol 122 patients undergoing elective cardiac or thoracic aortic surgery were evaluated preoperatively, 1 week, and 6 weeks postoperatively with a battery of neuropsychological tests. Seventy-two patients had routine cardiac surgery without DHCA (Group No DHCA), 50 patients had thoracic aortic surgery with varying periods of DHCA and were subdivided into those with 1-24 rain of DHCA (n = 26), and those with > 25 rain of DHCA (n = 24). The neuropsychological test battery consisted of 8 tests consolidated into 5 domains: attention, cognitive speed, memory, executive function, and fine motor function. Data were normalized to baseline values, and were analyzed using ANOVA, ANCOVA, and survival functions.

Results: Age was a significant predictor of impairment in memory, fine motor and executive function, therefore standardized scores were age-adjusted. Patients who could not be tested or had poor testing performance at 1 week postoperatively were more likely to perform poorly at 6-weeks (odds ratio 5.27, p<0.01). DHCA > 25 rain and to a lesser degree increasing age were significant predictors of impaired memory and motor function at 6 weeks postoperatively (Table 1), but not of attention, cognitive speed, or executive function. The decline (-0.48 ± 0.27) in memory function at 6 weeks (Table 2) roughly approximates a 20 point decrease in IQ. DHCA > 25 min (Odds ratio 4.0; p = 0.02) and increasing age (Odds ratio/5 years 1.23; p = 0.06) were determinants of prolonged hospital stay(≥21 days) (Table 1)

Conclusion: Prolonged DCHA (≥25 minutes) especially in older patients is associated with a subtle but important neuropsychological deficit in the domains of memory and fine motor function and also with prolonged hospital stay. The previously reported high incidence of temporary neurological dysfunction may be a clinical marker of this insidious neurological injury.

Table 1: Predictors of neuropsychological outcome (at 6 weeks) and hospital stay
	MEMORY	MOTOR	Hospital Stay (>21 days)
	Odds Ratio p	Odds Ratio p	Odds Ratio p
DHCA >25 min	3.52 0.04	7.25 0.004	4.01 0.02
AGE (/5 year incr.)	1.14 0.16	1.21 0.08	1.23 0.06

Table 2: Mean memory performance and change from baseline (age-adjusted Standard scores ± SEM)
	Baseline	6 Weeks Postop	Change
No DHCA	0.04 ±0.1 6	0.42 ±0.1 8	0.38
1-24 min DHCA	0.01 ±0.20	0.49 ± 0.24	0.48
> 25 min DHCA	-0.04 ± 0.24	-0.48 ±0.27*	-0.44*
* p < 0.05 compared with no DHCA at baseline, 6 weeks post-op, and change from baseline

11:25 a.m. C. WALTON LILLEHEI RESIDENT FORUM AWARD PRESENTATION

11:30 a.m. ADDRESS BY HONORED SPEAKER

A Practical Affair

Professor Ken Taylor, M.D., FRCS, London, England

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

12:10 p.m. CARDIOTHORACIC RESIDENT'S LUNCHEON

*By invitation