TUESDAY MORNING, APRIL 26, 1994

7:00 a.m. FORUM SESSION I - Grand Ballroom

Moderators: Andrew S. Wechsler, M.D.

Richard D. Weisel, M.D.

F1. MOLECULAR CARDIOMYOPLASTY - CARDIAC GENE THERAPY

Stanley K.C. Tarn, M.D.*, Wei Gu, Ph.D.* and Bernardo Nadal-Ginard, M.D., Ph.D.*

Cambridge and Boston, Massachusetts

Cardiomyoplasty, an operation consisting of synchronously paced skeletal muscle, commonly the latissimis dorsi, wrapped around the heart, offers a potential treatment of chronic heart failure, by using the power derived from skeletal muscle. Another strategy to treat heart failure is to provide additional endogenous power, i.e. more cardiac myocytes. Cardiac myocytes become terminally differentiated and lose their ability to undergo mitosis soon after the newborn period. As yet, no cardiac myogenic transcription factor has been identified. However, skeletal myogenic factors, the MyoD family of basic helix-loop-helix proteins, have been well-described. Studies have demonstrated that these basic helix-loop-helix proteins can function as master genes for induction of the skeletal muscle expression program, i.e. transfection of the MyoD gene has been shown to mediate the conversion of mouse C3H10T1/2 fibroblasts into stable myoblasts, which could be induced to undergo terminal differentiation into myotubes. Since essentially the same structural contractile apparatus is shared by both cardiac and skeletal myocytes, conversion of cardiac fibroblasts, for example those in a scar following myocardial infarction, into potentially functional skeletal myocytes may be of benefit for the treatment of heart failure. Sense and anti-sense MyoD genes were cloned into retroviral vectors carrying G418 resistance. Primary cardiac fibroblasts, freshly isolated from newborn rats, were infected with virus carrying sense or anti-sense MyoD gene. Ten days post-infection and post-G418 selection, expression of MyoD protein was demonstrated in 95% of cells infected with sense MyoD virus by intense nuclear immunostaining, using a MyoD polyclonal antibody. In contrast, none of the cells infected with anti-sense MyoD virus showed staining. Upon withdrawal of growth factors, 95% of MyoD positive cells became elongated and, in the presence of appropriate cell density, fused to form multi-nucleated cells, morphologically similar to striated muscle cell. Spontaneous contractile movements were noted in 10% of the cells. Expression of a myogenic differentiation marker, myosin heavy chain, in 95% of these elongated cells were detected by intense cytoplasmic immunostaining, using a myosin heavy chain monoclonal antibody. In contrast, MyoD negative cells remained unchanged. In summary, cardiac fibroblasts were able to be converted into bona fide potentially functional skeletal myocytes as shown by definitive morphologic and biochemical changes. Further studies are needed to explore this unique strategy to treat heart failure.

*By invitation

F2. NITROGLYCERIN MAINTAINS GRAFT VASCULAR HOMEOSTASIS AND ENHANCES PRESERVATION IN AN ORTHOTOPIC RAT LUNG TRANSPLANT MODEL

Yoshifumi Naka, M.D., Ph.D.*, Nepal C. Chowdhury, M.D.*, Mehmet C. Oz, M.D.*t, Robert E. Michler, M.D.*, Osvaldo J. Yano, M.D.*, Craig R. Smith, M.D., David M. Stern, M.D.* and David J. Pinsky, M.D.*

New York, New York

Pulmonary dysfunction following lung transplantation often occurs for poorly understood reasons. Because our pilot studies show that endothelial cells exposed to hypoxia and reoxygenation have undetectable nitric oxide (NO) levels due to the quenching effect of oxygen free radicals, and NO is vasodilatory, antithrombotic and prevents neutrophil adherence, we hypothesized that the NO donor nitroglycerin (NTG) might enhance lung preservation for transplantation. We used a new orthotopic rat lung transplant model in which the native PA is ligated so that hemodynamic assessment and survival depends exclusively upon the transplanted lung. The left lung was harvested from 22 rats in double blinded experiments, flushed (pressure < 20 mm Hg) with either lactated Ringer's (LR, n=11) or LR+NTG O.lmg/ml (LR+NTG, n=11), preserved for 4 hrs at 4°C, followed by orthotopic transplantation. Baseline measurements immediately upon reperfusion included arterial pO₂ (mm Hg), pulmonary artery flow (PAF, ml/min), mean PA pressure (MPAP, mm Hg), and left atrial pressure (LAP, mm Hg). After ligation of the right PA, these measurements were recorded simultaneously every 5 min for 30 minutes or until recipient death, after which myeloperoxidase assays (MPO, ∆ABS₄₆₀ nm/min) were performed to quantify neutrophil deposition. Pulmonary vascular resistance (PVR, Woods units x 1000) was calculated as (MPAP-LAP)/PAF. After 4 hrs preservation, 0% of the LR rats survived 30 min, whereas 64% of the NTG rats survived (p<0.01). Hemodynamic measurements (± SEM) recorded at the final time at which the recipient was alive are shown below (*=p<0.05).

	PAF	MPAP	PVR	p02	MPO
LR	0.4 ± 0.1	10.5 + 0.7	5.7 + 2 1	130 ± 39	2.6 + 0.3
LR+NTG	6.4 + 2.2*	17.4 ± 3.0*	1.1 ± 0.2*	339 + 66*	1.7 ± 0.3*

In similar experiments, nitroglycerin also significantly lengthens the effective preservation offered by Euro-Collins to over 8 hours. Supplementation of preservation solutions with NTG enhances blood flow, reduces PVR, diminishes leukostasis in the transplanted lung, and improves recipient survival. Nitroglycerin may be beneficial in clinical lung transplantation.

†1994-96 Research Scholar

*By invitation

F3. IMMEDIATE EARLY GENE EXPRESSION IN HUMAN SAPHENOUS VEINS HARVESTED DURING CORONARY ARTERY BYPASS SURGERY

Richard A. Moggio, M.D., Jia-Zhen Ding, M.D.*, Carolyn J. Smith, Ph.D.*, Robert R. Tota, M.D.*, Michael B. Stemmerman, M.D.* and George E. Reed, M.D.

Valhalla, New York

Saphenous vein graft (SVG) occlusion is a common late complication of coronary bypass (CABG). Intimal smooth muscle cell hyperplasia (SMC) is a component of this pathobiology, but the underlying molecular events are poorly understood. Immediate early genes (IEG) are activated shortly after growth stimulation and subserve cellular functions which may contribute to intimal SMC accumulation. In the present study, human SVG were harvested with minimal manipulation during CABG and processed for isolation of total RNA to examine changes in IEG mRNA expression by Northern blotting techniques. Thirty SVG were incubated at 4°C in Dulbecco's Modified Eagle media from 30 minutes to 6 hours. The mRNAs for c-fos and c-myc were weak or undetectable in controls, but were increased (>10 times controls) within one hour (c-fos). and persisted for at least 6 hours (c-myc)after harvest. c-fos is a transcription factor which can activate genes that affect cell growth and differentiation, c-myc regulates proliferation in various cell types. Anti-sense oligonucleotides have targeted c-myc mRNA and related IEGs, and have been shown to prevent intimal SMC hyperplasia in animal models of vascular injury.

Our results demonstrate, for the first time in human vascular tissue, incipient IEG induction. This information may lead to molecular therapies to control SVG disease, restenosis following angioplasty and atherosclerosis in general.

*By invitation

F4. GENE THERAPY USING ADENOVIRAL VECTOR TRANSFER OF THE HSV-THYMIDINE KINASE GENE AS AN APPROACH FOR TREATMENT OF HUMAN MALIGNANT MESOTHELIOMA

W. Roy Smythe, M.D.*, Harry C. Hwang, B.A.*, Kunjlata M. Amin, Ph.D.*, James M. Wilson, M.D., Ph.D.*, Steven M. Albelda, M.D.* and Larry R. Kaiser, M.D.

Philadelphia, Pennsylvania

Malignant pleural mesothelioma is a neoplasm that is unresponsive to conventional cell and tissue-level toxic chemotherapies, however, gene therapy offers potential to treat such tumors at a subcellular level with minimal toxicity to normal cells. Cells expressing the herpes simplex thymidine kinase gene (HSVtk) may be killed when exposed to ganciclovir (GCV) due to production of a toxic nucleotide analog that interferes with cell replication. We theorized that recombinant adenoviral vectors could transfer HSVtk to mesothelioma and effect cell killing both in vitro and in vivo after exposure to GCV.

Two human malignant mesothelioma cell lines were infected by adenovirus vectors carrying a Rous Sarcoma Virus promoter and either the HSVtk gene (Ad.RSV.tk) or a non-therapeutic marker gene (Ad.RSV.lacZ - gene for β-galactosidase) at a multiplicity of infection of 100 viral particles/cell. Cells were then exposed to varying concentrations of GCV in media for 4 days, with viable cell growth assessed by colorimetric assay. GCV sensitivity, expressed concentration necessary to eliminate 50% of viable cells (IC50) is shown below.

Table 1. Mcsothclioma / GCV IC50
Cell Type	AclRSVtk	AdRSVIacZ	uninfected
1-45	<0.1 mM	> 1000 mM	>750 mM
REN	< 1.0 mM	> 1000 mM	>1000mM

2 x 10⁷ 1-45 cells infected in vitro with either Ad.RSV.tk or Ad.RSV.lacZ were then injected into subcutaneous flank tissues (R=lacZ, L=HSVtk) of three Severe Combined Immune Deficient mice. After three days, animals had developed bilateral tumors of 6-8 mm diameter and began treatment with 5 mg of intraperitoneal GCV/day. By day ten, left flank tumors in all animals (HSVtk side) were undetectable, with continued growth of tumors derived from Ad.RSV.lacZ infected cells.

In conclusion, we demonstrate that recombinant adenoviral vectors can transfer a therapeutic HSVtk prodrug gene to human mesothelioma, effecting substantial acute in vitro cellular killing and complete in vivo tumor regression with exposure to non toxic doses of the drug GCV.

*By invitation

F5. INTRAUTERINE CREATION OF INCREASED PULMONARY BLOOD FLOW IN LAMBS: POSTNATAL PULMONARY HYPERTENSION AND EARLY PULMONARY ENDOTHELIAL DYSFUNCTION

V. Mohan Reddy, M.D.*, Jackson Wong, M.D.*, John L. Liddicoat, M.D.*, Frank L. Hanley, M.D.* and Jeffrey R. Fineman, M.D.*

San Francisco, California

Sponsored by: Benson B. Roe, M.D., San Francisco, California

Pulmonary hypertension is a common accompaniment of congenital heart disease with increased pulmonary blood flow, and remains a major source of morbidity and mortality in the peri- and post-operative period. Previous models developed to study this problem have entailed placement of aorta to pulmonary shunts postnatally; after pulmonary vascular resistance has fallen and a period of lung growth and development has occurred. To establish a model which mimics the clinical setting of increased pulmonary blood flow during the transition circulation and immediately after birth, we placed gore-tex vascular grafts between the aorta and main pulmonary artery in 8 late gestation fetal sheep.

Recent evidence suggests that endothelial injury secondary to increased pulmonary blood flow disrupts the normal regulatory mechanisms provided by the vascular endothelial cells, and may contribute to the development of pulmonary hypertension and increased vascular reactivity. To investigate this hypothesis, the responses to ATP (an endothelium-dependent vasodilator), sodium nitroprusside (SNP, an endothelium-independent vasodilator), endothelin-1 (ET-1, a vasoactive polypeptide produced by vascular endothelium), and alveolar hypoxia (8.0% oxygen) in these shunted lambs (at 4 weeks of age) were compared to historical (1-2 week old) controls.

Six fetuses had 4.0-6.0 mm grafts placed between the bovine trunk and the main pulmonary artery. Four weeks after spontaneous delivery, mean pulmonary arterial pressure (PAP) was 24.7 ± 3.9 mmHg (controls = 13.0 ± 3.0 mmHg). The ratio of PAP to systemic arterial pressure (SAP) was 0.27 ± 0.09. The ratio of pulmonary to systemic blood flow (Qp/Qs) was 2.1:1 ± 0.7. Two fetuses had an 8.0 mm graft placed between the ascending aorta and the main pulmonary artery. At 4 weeks of age, PAP was 58.5 ± 7.7 mmHg, PAP/SAP was 0.95 ± 0.05, Qp/Qs was 1.6:1 ± 0.2, and pulmonary vascular resistance was 3.1 ± 0.3 mmHg min-kg L^-1(control = 0.65). The percent changes in PAP in response to the vasoactive stimuli are shown in the table below.

	PERCENT CHANGE IN PAP
	ATP	SNP	ET-1	HYPOXIA
SHUNTED	-0.9% ± 1 .9*	-29.9% ± 8.8	+31. 8% ± 14.4*	+60.4% ± 32.7*
CONTROLS	-24.7% ± 2.5†	-33.5% ± 15.4†	-24.1% ± 7.5†	+35.6% ± 17.3
* p<0.05, vs. Controls Value by the unpaired t-test † Value obtained during pulmonary hypertension induced by U4661 9 (a thromboxane A2 mimic).

This preparation is the first model of pulmonary hypertension secondary to congenital increased pulmonary blood flow, and is a useful tool to study the development of pulmonary hypertension and increased vascular reactivity secondary to congenital heart disease. Increased pulmonary blood flow produces early selective endothelial dysfunction (as manifested by impaired endothelium-dependent vasodilation and ET-1-induced vasoconstriction) and increased vascular reactivity to hypoxia. A better understanding of the development of pulmonary hypertension secondary to congenital heart disease will optimize the timing of surgical repair and improve the peri- and postoperative management.

*By invitation

F6. MARKED ENHANCEMENT IN MYOCARDIAL FUNCTION RESULTING FROM OVEREXPRESSION OF A HUMAN P-ADRENERGIC RECEPTOR GENE

Carmelo A. Milano, M.D.*, Lee F. Allen, M.D., Ph.D.*, Paul C. Dolbert, Ph.D.*, David Johnson, Ph.D.*, Howard A. Rockman, M.D.*, Richard Bond, Ph.D.* and Robert J. Lefkowitz, M.D.*

Durham, North Carolina and Houston, Texas

Molecular biological approaches have affected significant advances in the understanding and treatment of disease; however, it has been difficult to apply this technology to the study of myocardial function. Using transgenic animals, this study examines whether overexpression of a β-adrenergic receptor (β-ADR) gene can improve baseline myocardial function.

Transgenic mice were created using a fusion gene composed of the coding sequence for the human β₂-ADR flanked at its 5' end by the alpha-myosin heavy chain promoter-enhancer sequence. Northern blot analysis of total RNA from six muscle containing tissues demonstrated intense, cardiac specific expression. Three lines of transgenic mice overexpressing the receptor were established. Radioligand binding assays performed on myocardial membrane fractions from the line with highest expression, quantitated total β-ADR as 35.4 ± 13.3 pmol/mg protein in the transgenics vs. 0.2 ± 0.09 pmol/mg in controls (n=4; 175 fold increase). Immunohistochemical staining with a human β2-ADR specific antibody revealed receptor protein expression on the sarcolemma in all four cardiac chambers. Adenylyl cyclase activity was determined by incubating membrane fractions with [³²P]-ATP and quantitating the rate of [³²P]-cAMP formation. In four independent experiments, both basal and isoproterenol stimulated cyclase activity were increased in the transgenics vs. paired littermate controls: basal - 290 ± 104 pmol/min/mg protein vs. 160 ± 27 (p<0.05, paired T-test); isoproterenol stimulated - 495 ± 123 vs. 356 ± 112 (p<0.05). In addition, the isoproterenol dose response curve was shifted markedly to the left in the transgenics relative to controls, indicating greater isoproterenol sensitivity. Myocardial function was assessed first in the atrial tissue where gene expression was greatest; isometric tension development was measured in both right and left atria hung in organ perfusion baths. To simulate in vivo conditions, right atria were allowed to contract at their intrinsic rate, and left atria were paced at the rate of their corresponding right atrium; basal and maximal isoproterenol-stimulated tensions were recorded. Comparing 6 transgenic right atria to 7 controls, basal transgenic tension was increased 170.8 ± 24.9 mg vs. 104.2 ± 16.2, (p<0.05, student's T-test). Comparing 5 transgenic left atria to 8 controls, basal transgenic tension was increased 3 fold: 239 ± 49.3 mg vs. 81.9 ± 24.1 (p<0.005). In addition, basal left atrial tension in transgenics equaled or exceeded maximally stimulated tension in the controls: 239 ± 49.8 mg vs. 173.1 ± 13 (p=0.143). Histologic examination of the hearts of 2 month old animals with collagen-specific stains demonstrated no changes in collagen content relative to controls; there was no evidence of fibrosis or myocyte necrosis. Heart/body weight ratios were not statistically different in the transgenics relative to nontransgenic littermate controls (n=6 pairs).

In conclusion, the alpha-myosin heavy chain promoter-enhancer sequence affects intense cardiac specific gene expression. Overexpression of (β-ADRs results in a marked enhancement of both basal biochemical and physiologic parameters of myocardial function. The intrinsic myocardial dysfunction and reduced β-ADR-mediated adenylyl cyclase activity common to many forms of heart failure, suggest a potential therapeutic role for (β-ADR Overexpression. As methods for in vivo gene transfer develop, targeted Overexpression of specific genes in the myocardium may serve as an important adjunct to current surgical or medical treatments.

*By invitation

F7. RETROGRADE CEREBRAL PERFUSION FOR DISLODGEMENT OF SOLID CEREBRAL EMBOLI

Alon S. Aharon, M.D.*, Hillel Laks, M.D., Jeffrey Pearl, M.D.*, Ehud Rudis, M.D.*, Lester Permut, M.D.*, John Frazee, M.D.*, Eli Ziv, B.S.*, Gary Mathern, M.D.*, Anthony Verity, M.D.*, Davis Drinkwater, M.D.*, Eli Kaczer, B.S.* and Paul Chang, B.S.*

Los Angeles, California

A primate model was created in order to investigate the efficacy of retro-perfusing ischemic brain following embolization of atherosclerotic debris during cardiac surgery. Retrograde cerebral perfusion via the superior vena cava (SVC) with oxygenated whole blood and moderate hypothermia was used as a method of washing out embolized atherosclerotic plaque and supporting ischemic brain in the baboon. Animals were divided into a control group (n=4) which underwent selective internal carotid artery (ICA) embolization of atherosclerotic plaque and a treated group (n=4) which underwent selective ICA embolization followed by cerebral retroperfusion via the SVC. Animals were anesthetized with Forane inhalation anesthetic and paralyzed with a non-depolarizing agent. Cardiopulmonary bypass (CPB) at blood flows of 100 ml·kg^-1·min^-l and mean arterial pressure above 70 mm Hg was initiated with bicaval cannulation and moderate systemic hypothermia to 28 degrees C. Both groups underwent cold blood cardioplegic arrest and selective ICA embolization using 10 mg of human atherosclerotic debris ranging in size from 250 to 500 microns. Following embolization control animals were rewarmed and weaned from CPB. Experimental animals underwent an additional 5 minute period of cerebral retroperfusion via the SVC with cold oxygenated blood at a mean perfusion pressure of 40 mm Hg before rewarming and weaning from CPB. Fourteen channel electroencephalogram (EEG) performed immediately after termination of CPB revealed signs of severe central nervous system (CMS) injury in all control animals which included: non reactive backgrounds signifying brain death in 3 of 4 animals, persistent lateralizing epileptiform discharges in 3 of 4 animals and lateralized findings in 4 of 4 animals. In the retroperfusion group post-CPB EEG revealed no evidence of significant CNS injury and all animals remained cortically reactive. Somatosensory evoked potentials (SSEP) using bilateral median nerve stimulation obtained in the last 5 animals revealed loss of thalamo-cortical peaks in 3 of 3 control animals and retention of thalamo-cortical peaks in 2 of 2 retroperfused animals. SSEP demonstrated a decrease in cortical reactivity 10-45 minutes after embolization in all animals with large areas of cerebral infarction. Microscopic analysis of formalin perfused fixed brains in animals sacrificed 36 to 48 hours after embolization revealed greater volumes of infarction in control animals when compared to animals treated with retroperfusion (1% vs. 5% via planimetry, 1% vs. 6% via computer digital analysis). Conclusions: (1) SSEP monitoring appears to correlate with early CNS injury and may be a valuable monitoring technique and (2) cerebral retroperfusion performed during CPB with moderate hypothermia can prevent or decrease the extent of CNS ischemic injury seen after embolization of atherosclerotic debris which may be an important cause of stroke after cardiac surgery.

*By invitation

F8. PLATELET FACTOR 4 - AN ALTERNATIVE TO PROTAMINE

Alvise F. Bernabei, M.D.*, Nicolas Gikakis*, Theodore E. Malone, Ph.D.*, Stefan Niewiarowski, M.D., Ph.D.* and L. Henry Edmunds, Jr., M.D.

Detroit, Michigan; Philadelphia, Pennsylvania and Cambridge, Massachusetts

Protamine sulfate (PS), used to inhibit heparin, often causes adverse reactions after cardiopulmonary bypass (CPB). Platelet factor 4 (PF4), a natural protein stored in platelet alpha granules, inhibits heparin. We compared the efficacy, safety and side effects of recombinant PF4 (rPF4) and PS inhibition of heparin in nine young female adult baboons (12-18 kg) with and without CPB.

In 12 trials each anesthetized, intubated baboon received 100 units/kg heparin IV; after 5-10 minutes, heparin was completely neutralized by bolus injection of 1 mgm/kg PS or 2 mgm/kg rPF4. Mean arterial, pulmonary arterial and pulmonary capillary wedge pressure decreased significantly (p>0.05) after PS only.

In 13 more trials, each baboon received 300 units/kg heparin and was perfused for 30 minutes at 50 ml/kg/min from right atrium to femoral artery using 8-10 Fr polyurethane cannulas, a roller pump, and an 0.8 M2 spiral coil oxygenator. Following decannulation, either 3 mgm/kg protamine or 6 mgm/kg rPF4 was injected. Neither drug caused significant (p>0.05) change in mean arterial, pulmonary arterial, central venous or wedge pressures or in cardiac output (thermodilution) 5 and 30 minutes after drug injection. Thrombin time and partial thromboplastin time remained significantly prolonged five minutes after injection of protamine, but returned to baseline at 30 minutes; both times were normal five minutes after rPF4. In contrast to rPF4, PS had an anticoagulant effect at high concentrations. After both drugs, template bleeding times were prolonged. Platelet count and responses to ADP did not differ between drugs. rPF4 increased activated complement 3 (C3a) from 91.3 ± 7.7 ng/ml to 150 ± 3.5 at 5 min; protamine increased C3a from 72 ± 4.2 to 99.4 ± 5.5 (p<0.05, between groups). All animals remain healthy.

We conclude that rPF4 neutralizes heparin faster than PS and has no anticoagulant effect at high concentration. Both drugs activate complement, but in baboons the small increases arc unimportant. Neither drug causes adverse side effects in baboons. The data support a decision for a clinical trial of rPF4.

*By invitation

F9. THE NOVEL EFFECTS OF 3, 5, 3' TRIIODOTHYRONINE UPON MYOCYTE CONTRACTILE FUNCTION AND (3-ADRENERGIC RESPONSIVENESS IN DILATED CARDIOMYOPATHY

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

Charleston, South Carolina

The number of patients undergoing cardiac surgery with chronic left ventricular (LV) dysfunction is increasing. These patients frequently require intensive inotropic support in the perioperative period. Recent clinical and experimental studies have suggested that 3, 5, 3' triiodothyronine (13) improves LV pump function. However, whether T₃ directly improves myocyte contractile function in cardiomyopathic disease is unknown. Accordingly, this study examined the direct effects of 73 upon isolated myocyte contractile function in cells obtained from control (n=6) pigs and pigs with tachycardia induced dilated cardiomyopathy (DCM; atrial pacing at 240 bpm for 3 weeks; n=6). Myocyte percent shortening (MYO-%) and myocyte velocity of shortening (MYO-VEL µm/s) were obtained at baseline and in the presence of T₃ (1-100 pM T₃). (*p<0.05 vs baseline, # p<0.05 vs control myocytes)

	Baseline		80pM T3		100pM T3
	MYO-%	MYO-VEL	MYO-%	MYO-VEL	MYO-%	MYO-VEL
Control (n=30)	4.4±0.1	43.8±1.5	6.0±0.3*	58.9±3.6*	6.1±0.3*	69.9±3.7*
DCM(n=14)	2.1±0.1	26.7+1.3	2.5±0.1*#	39.9±5.6*#	2.8±0.5*#	44.1±102*#

For both control and DCM groups, T₃ (80 and 100 µM) caused a significant increase in myocyte contractile function. Clinically, a common method of inotropic support is stimulation of the (3-adrenergic receptor system (PAR). Accordingly, a second series of experiments was performed in which control and DCM myocyte contractile function was examined with the (βAR agonist, isoproterenol (ISO=25nM) alone, and in myocytes preincubated with 80 and 100 pM T₃ to which isoproterenol was added. (+p<0.05 vs ISO)

	Isoproterenol		SOpM T3 + ISO		100pM T3 + ISO
	MYO-%	MYO-VEL	MYO-%	MYO-VEL	MYO-%	MYO-VEL
Control (n=30)	8.8±0.7	100.2±1.5	11.3±0.7+	152.9±11.0+	11.8±0.6+	155.1±8.5+
DCM(n=14)	4.9±0.5	73.7±9.1	5.6±0.9+	129.3±28.5+	5.7±1.4+	116.8±29.8+

For both the control and DCM groups, myocyte contractile function increased with preincubation of T₃ over isoproterenol values alone. Further, MYO-VEL increased by over 300% in DCM myocytes following preincubation with 100 pM T₃ followed by ISO compared to a 200% increase in control myocytes (p<0.05). The results from this study clearly demonstrated that T₃ directly augments myocyte contractile function in both control and DCM myocytes. In addition, T₃ significantly enhanced myocyte contractile function following βAR stimulation in cardiomyopathic myocytes. This study provides unique evidence to suggest that T3 may be a useful adjunct to conventional inotropic support in the setting of advanced left ventricular dysfunction.

*By invitation

9:00 a.m. SCIENTIFIC SESSIONS - Grand Ballroom

Moderators: Delos M. Cosgrove, M.D.

Douglas J. Mathisen, M.D.

13. INTERMEDIATE RESULTS FOLLOWING COMPLETE REPAIR OF TETRALOGY OF FALLOT IN NEONATES

Hani A. Hennein, M.D.*, Ralph S. Mosca, M.D.*, Gonzalo Urcelay, M.D.*, Dennis C. Crowley, M.D.* and Edward L. Bove, M.D.

Ann Arbor, Michigan

From July 1988 through September 1993, 30 neonates with symptomatic tetralogy of Fallot (TOP) underwent complete repair. Sixteen patients had uncomplicated TOP, 9 had TOP and pulmonary atresia (PA), 3 had TOP with nonconfluent pulmonary arteries (NCPA), and 2 had TOP, PA and NCPA. The median age at operation was 11 days (mean ± SEM, 12.6 ± 2.9 days) with a mean weight of 3.1 ± 0.1 kg (range, 1.5 - 4.4 kg). Preoperatively, 14 patients were receiving an infusion of prostaglandin, 13 were mechanically ventilated, and 6 required inotropic support. Right ventricular outflow obstruction (RVOTO) was managed by a limited transannular patch in 25 patients, infundibular muscle division with limited resection in 15, and insertion of an RV-PA valved aortic homograft conduit in 5 patients. Follow up was complete at a median interval of 24 months (range, 1-62 months). There were no hospital deaths and two late deaths resulting in one month, one year and five year actuarial survival rates of 100%, 93% and 93%, respectively. The hazard function for death had a rapidly declining single phase that approached zero by 6 months after surgery. Both deaths occurred in patients with TOP/PA who had undergone aortic homograft conduit RVOT reconstruction, making the only independent risk factor for death the use of a valved aortic homograft conduit (p<0.005). Eight patients required reoperation, resulting in one month, one year and five year freedom from reoperation rates of 100%, 93% and 66%, respectively. Indications for reoperation were branch left pulmonary artery stenosis in 5 patients, RVOTO in 2 patients, and severe pulmonary insufficiency in 1 patient. Independent risk factors for reoperation included an intraoperative RV/LV pressure ratio of 0.75 or greater (p=0.04), Doppler residual left pulmonary artery stenosis of ≥15 mmHg, or Doppler RVOT gradient of ≥40 mmHg at hospital discharge (p=0.002 and 0.02, respectively). This series demonstrates the safety of early complete repair of symptomatic tetralogy of Fallot in neonates. It also emphasizes the importance of relieving all sources of RVOTO at the initial operation, particularly that located at the ductus insertion site, which may be difficult to diagnose in the neonate before ductal closure occurs. The safety and efficacy of valved aortic homograft conduits in neonates requires further investigation.

*By invitation

14. OPTIMIZING SELECTION OF PATIENTS FOR MAJOR LUNG RESECTION

Mark K. Ferguson, M.D.†, Laurie B. Reeder, M.D.* and Rosemarie Mick, M.S.*

Chicago, Illinois

Background: Diffusing capacity (DLCO) correlates with mortality and pulmonary morbidity following lung resection. It is not known whether a normal DLCO permits safe resection in patients with marginal spirometric values, or whether normal spirometric values negate the adverse effects of a low DLCO. The purposes of this study were: 1) to determine the best predictors of morbidity and mortality; and 2) to assess whether interactions exist between DLCO and spirometry that help estimate outcome after major lung resection.

Methods: We performed a retrospective analysis of 376 patients who underwent lung resection from 1980 to 1992 (222 men, 154 women; mean age 60.1 years). 307 had lung cancer (Stage I:127; II:55; IIIa:120) and 69 had other disease. 284 underwent lobectomy/bilobectomy and 92 had pneumonectomy. We assessed the relationship of 21 preoperative variables to 20 postoperative events classified as pulmonary (PULM) or (CARD) complications, overall nonfatal morbidity (MORE), and operative mortality (MORT).

Results: The best single predictor of complications was the predicted postoperative DLCO (ppoDLCO%; calculated by multiplying the percentage of unresected lung segments by the measured preoperative DLCO expressed as a percent of predicted; p<.004 for each outcome by univariate analysis). The incidence of complications was inversely related to ppoDLCO% (Cochran-Armitage trent test):

		ppoDLCO%→	<40	40-50	50-60	60-70	>70	p
	PULM	(53/279)	33.3	16.7	32.7	16.9	8.3	.001
Incidence	CARD	(52/260)	31.2	27.3	19.1	18.5	14.9	.005
(%)	MORB	(121/270)	55.9	47.1	61.5	35.5	36.3	.003
	MORT	(19/289)	19.4	13.2	5.3	1.6	3.1	.0003

Stepwise logistic regression analysis based on 9 significant variables (age, sex, NYHA class, prior MI, FVC%, DLCO%, FEV₁%, ppoFEV₁%, ppoDLCO%) identified only ppoDLCO% and age as predictors of PULM, MORB and MORT, and these with prior MI predicted CARD (model significance p<.001 by χ² for each complication). There was no interaction between spirometry and ppoDLCO% in predicting complications. Eliminating patients with ppoDLCO% <50 theoretically would have reduced the mortality by >50% (from 6.5% to 3.2%).

Conclusion: DLCO measures subtle but important changes in lung function that often are undetected by spirometry. It independently and strongly predicts the risk of complications and mortality following lung resection, and its routine use is invaluable in the preoperative assessment of lung resection candidates.

†1986-88 Research Scholar

*By invitation

15. ESTIMATION OF PATIENT-SPECIFIC RISK OF PROSTHETIC VALVE REOPERATION WITH REFERENCE TO PROPHYLACTIC VALVE REPLACEMENT

Jeffrey M. Piehler, M.D., Eugene H. Blackstone, M.D., Kent R. Bailey, Ph.D.*, Michael E. Sullivan, M.D.*, James R. Pluth, M.D, Noel S. Weiss, M.D, Dr.P.H.* and Ronald S. Brookmeyer, Ph.D.*

Kansas City, Missouri; Birmingham, Alabama; Rochester, Minnesota; Portland, Oregon; Scottsdale, Arizona; Seattle, Washington and Baltimore, Maryland

One of the major variables in the decision to prophylactically replace a prosthetic heart valve is a patient-specific estimate of operative risk. To date no study has provided the means to make this estimation for both low and high risk subgroups. Consequently, a cooperative review of 2246 prosthetic valve reoperations performed between 1963 and 1992 at three experienced academic institutions was undertaken. The records were subjected to multivariable logistic regression analyses of 53 variables with potential influence upon hospital mortality. Those variables which would be known preoperatively and which were found to have the strongest correlation with mortality (Table) were incorporated into a predictive risk equation which was internally validated. Predicted risks varied widely from less than 1% to greater than 99%. No institutional or temporal risk factors were identified.

Incremental Risk Factor	p-value
Demographic
Age	<.0001
Weight	.0008
Left Ventricular Function and Secondary Conditions
NYHA Class (I-V)	<.0001
Hemodynamic Status	<.0001
Tricuspid Incompetance	<0001
Extensiveness of Valvular Heart Disease
Multiple Valve Disease	.002
Previous Cardiac Surgery
Number of Previous Open Heart Operations	.001
Coexisting Morbid Conditions
Prosthetic Valve Endocarditis	<0001
Renal Failure	.01
Concomitant Procedures
Repair of Ascending Aortic Aneurysm	<0001
Coronary Artery Bypass Grafting	.009
Left Ventricular Aneurysmectomy	.005

With limits placed upon contributing risk factors, the predictive risk equation can be utilized to construct nomograms for estimating operative risk (Figure). Nevertheless, solving the equation for an individual patient is relatively straightforward and markedly more specific. For example, estimated risk of a first elective single valve reoperation in a 55-year old woman who is NYHA Class I, without coexisting morbid conditions or concomitant procedures is only 1.5% (90% confidence limits 1.2%-1.9%). In contrast, estimated risk for the same operation in a 68-year old woman, Class III, with elevated creatinine, tricuspid incompetence, and the need for coronary bypass is 43% (33%-54%). We conclude that hospital mortality for prosthetic valve reoperation can be accurately predicted and is generally low for most uncomplicated patients. Such data has applicability to analysis of prophylactic valve replacement and other clinical situations.

*By invitation

10:45 a.m. SCIENTIFIC SESSION - Grand Ballroom

Moderators: Aldo R. Castaneda, M.D.

James L. Cox, M.D.

16. FACTORS ASSOCIATED WITH EARLY AND LATE RISK AFTER FONTAN OPERATIONS

John E. Mayer, Jr., M.D., Thomas Gentles, M.D.*, John Kupferschmid, M.D.*, Gil Wernovsky, M.D.*, Richard A. Jonas, M.D. and Aldo R. Castaneda, M.D., Ph.D.

Boston, Massachusetts

Early survival after Fontan operations for functional single ventricle (including tricuspid atresia) show continued improvement despite expanded selection criteria. Few reports provide information on factors influencing longer term outcome. Among 500 consecutive Fontan operations (1973-1991) 82 early and 36 late failures (death, transplant or takedown) occurred. In 1992-93 we obtained follow-up on 392 of the 418 early survivors. Separate analyses were carried out for early and late failure. Univariant analysis identified potentially risk factors (p>0.1) which were entered into logistic regression analyses. Risk factors for early failure (all with p<.05) were age <4 yrs, any history of PA hypertension, presence of a common AV valve, aortic saturation <80%, PA pressure >15mmHg, pulmonary arteriolar resistance >2 Wood units, pulmonary artery distortion, tricuspid valve as the systemic AV valve, and use of right atrial appendage or free wall in Fontan pathway. Fenestration reduced early risk. However, the only risk factors for late failure were atriopulmonary (vs cavopulmonary) anastomoses, presence of a pacemaker, and aortic cross clamp time >55 minutes. The late survival curve shows a continuing risk out to at least 15 years. Thus, once patients survive the Fontan operation, morphologic and physiologic characteristics have little impact on late survival, but surgical technique variables continue to exert an effect.

*By invitation

17. HEMODYNAMIC AND PHYSIOLOGIC CHANGES DURING IMPLANT ABLE LVAD SUPPORT

Patrick M. McCarthy, M.D.*, Karen B. James, M.D.*, Emmanuel L. Bravo, M.D.*, Robert M. Savage, M.D.*, Rita Vargo, R.N.*, Shelly K. Sapp, M.S.*, Marlene Goormastic, M.P.H.* and James D. Thomas, M.D.*

Cleveland, Ohio

Sponsored by: Delos M. Cosgrove, M.D., Cleveland, Ohio

To evaluate hemodynamic effectiveness and changes in patient (pt) physiology on the HeartMate® left ventricular assist device (LVAD), we studied 19 moribund bridge to heart transplant pts (35 to 62 years old. mean 50 years). All pts were on inotropes pre-LVAD, 16 (84%) were on a balloon pump, and 3 (16%) were on heparin-coated ECMO for circulatory support. Three pts died, all had right ventricular (RV) dysfunction and developed multiple organ failure. Three pts are rehabilitated and are awaiting a donor. Of the 16 survivors (84%), 13 have been transplanted (Tx), and post-Tx survival is 100%. Duration of LVAD support averaged 66 days (22 to 101 days). Pts on support over 30 days were NYHA class I before Tx. There were no thromboembolic events with over 1100 patient days of support. Only aspirin and persantine were administered. Significant hemodynamic improvement occurred:

	Pre-LVAD (n=19) (mean ± SD)	Pre-Tx (n=13) (mean ± SD)	p value
Cardiac index (L/inin/m2)	1.6 ± 0.2	3.2* ±0.9	.0002
Left atrial pressure (mmHg)	22.9 ± 9.5	8.0 ± 5.5	.034
RV ejection fraction	19.8 ± 11.3	40.8 ± 8.9	.005
Pulmonary vascular resistance (Wood units)	5. 2 ±2.6	2.0 ± 0.8	.0001
* mean "pump index during support

Bilirubin, blood urea nitrogen, plasma renin activity, angiotensin II, arginine vasopressin, plasma epinephrine, and plasma norepinephrine all decreased to normal, or near-normal, pre-Tx (all values p<.001 percent change from baseline pre-LVAD). Atrial natriuretic peptide levels decreased (326 ± 301 pg/ml to 185 ± 89 pg/ml) but did not return to normal (26-33 pg/ml).

We conclude that: 1) implantable LVAD support improved hemodynamic function in 84% of pts; 2) NYHA class improved, as well as subsystem function and the neurohormonal axis; 3) chronic LVAD support with a low risk of thromboemboli is attainable.

*By invitation

11:25 a.m. ADDRESS BY HONORED SPEAKER

A Thoracic Tale of Two Cities

Rodolfo Herrera-Llerandi, M.D., Guatemala City, Guatemala

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

12:10 p.m. CARDIOTHORACIC RESIDENTS' LUNCHEON -

Petite Trianon