TUESDAY MORNING, APRIL 30, 1996

7:00 a.m. FORUM SESSION I - CARDIAC SURGERY

Room 6A/B, San Diego Convention Center

Moderators: Edward D. Verrier, M.D.

James K. Kirklin, M.D.

F1. EFFICIENT TRANSFER OF OLIGONUCLEOTIDE AND PLASMID DNA INTO WHOLE HEART THROUGH CORONARY ARTERY.

Yoshiki Sawa, M.D.*, Keishi Kadoba, M.D.*, Kazuhiro Taniguchi, M.D.*, Hong-zhi Bai, M.D.*, Ken Suzuki, M.D.*, Yasufumi Kaneda, M.D.* and Hikaru Matsuda, M.D.*

Osaka, Japan

Sponsored by: Yasunaru Kawashima, Osaka, Japan

At the time of harvesting donor heart, it may be possible to perform gene transfection to provide the myocardium protection from the ischemic injury and/or change the alloreactivity after transplantation. However, several of the current techniques for transfer of both oligonucleotide and plasmid DNA into the myocardium are impaired by low efficiency and toxicity. To improve gene transfer techniques especially into whole heart, a gene transfer method involving liposome with viral envelope (HVJ-liposome) was assayed as an alternative. In this study, in vivo gene transfection of FITC labeled oligonucleotide (F-ODN) and cDNA of p-galactosidase (P-gal) and Mn-SOD was examined using HVJ-liposome (H group) or cationic liposome (L group). In H group, F-ODN or cDNA of P-gal or MnSOD were complexed with liposomes, DNA binding nuclear protein (HMG-1) and the viral protein coat of HVJ. After the harvest of donor rat hearts arrested by cardioplegia, coronary artery was infused with liposome-gene complex through aortic cannula during cardioplegia arrest. Then the hearts were transplanted (anoxic time: 30 min.) into the abdomen of the recipient rat of the same strain and all hearts were sacrificed after 3 days of transplantation. All hearts showed no rejection at the time of sacrifice. FITC was detected in the nuclei of more than 80% of the myocytes (82 ± 17%) in H group than in L group (7 ± 5%). Intensity of FITC was significantly higher in H group (939 ± 112 F.I.) than L group (166 ± 78 F.I.). P-gal was expressed in the cytosol of more than 70% of the myocytes (71 ± 14%). After 3 days of gene transfection, the hearts transfected with Mn-SOD (S) showed significantly higher percentage of recoveries of LVDP (S vs C, 86 ± 3 vs 54 ± 12%) and coronary flow (98 ± 2 vs 66 ± 12%) than did the control hearts (C) when exposed to ischemia (30 min., 37°C) and reperfusion (30 min., 37°C) with Langendorff apparatus. These results clearly demonstrated that donor hearts were transfected with FITC-oligonucleotide and p-galactosidase gene in the whole layer of myocardium by coronary infusion of HVJ-liposome during cardioplegia arrest at the time of harvest. And the heart transfected with Mn-SOD showed significant improvement of tolerance against ischemia-reperfusion injury. Our method appears to be a novel in vivo gene transfer technique for the heart to be subjected to ischemia, and may provide a new tool for research and therapy of heart transplantation.

*By invitation

F2. TIROFIBAN PROVIDES "PLATELET ANESTHESIA" DURING CARDIOPULMONARY BYPASS IN BABOONS.

Yugi Hiramatsu, M.D.*, Nicolas Gikakis, B.S.E.*, Harry L. Anderson, M.D.*, Joseph H. Gorman, M.D.*, Robert J. Gould, M.D., Ph.D.*, Stefan Niewiarowski, M.D.* and L. Henry Edmunds, Jr., M.D.

Philadelphia, Pennsylvania

Tirofiban is a reversible, non-protein inhibitor of platelet GPIIb/IIIa receptors that is in Phase III clinical trials for some applications. We tested the efficacy of tirofiban to preserve platelet numbers and function and to shorten postoperative bleeding times in a baboon model of cardiopulmonary bypass (CPB).

Three groups of baboons were studied: control (n=9); low dose tirofiban (0.1 ug/kg/min, n=7); and high dose tirofiban (0.3 ug/kg/min, n=7). Tirofiban was infused 60 min before and during CPB. After heparin (300 n/kg), animals were perfused for 60 min at 50 ml/kg/min at 37°C using a bubble oxygenator (to increase platelet activation) and peripheral cannulation. Hemodynamics, platelet count, aggregation to ADP, release of PTG and template bleeding times (TBT) were measured before tirofiban infusion; before heparin; after heparin before CPB; after five and 55 min of CPB; after protamine (3 mg/kg); and 60 min after protamine. TBT were also measured 120 and 180 min after protamine. Platelet GPIIIa antigen was measured in Triton X-100 washes of the perfusion circuit after CPB.

High dose tirofiban completely prevented platelet loss and release of βTG. ADP aggregation was suppressed during CPB, but returned to control values 60 min after protamine. These measurements did not significantly differ from control with low dose tirofiban but did with the high dose. TBT in control animals was 11.6 and 10.5 min at 120 and 180 min after protamine. Both doses of tirofiban significantly (p < 0.004 - 0.02) shortened TBT at the same time points to 7 (120 min) and 5 min (180 min) after protamine. Surface GPIIIa antigen did not significantly differ between groups. Neither dose of tirofiban altered hemodynamics.

High dose tirofiban completely preserves platelet numbers and function during CPB in baboons and significantly accelerates restoration of normal template bleeding times.

*By invitation

F3. CORRELATION OF FUNCTIONAL PROGNOSIS WITH MYOCARDIAL MORPHOLOGY AND METABOLISM IN PATIENTS UNDERGOING CORONARY BYPASS SURGERY.

Christophe Depre, M.D.*, Jean-Louis Vanoverschelde, M.D., Ph.D.*, Bernard Gerber, M.D.*, Marcel Borgers, M.D., Ph.D.*, Jacques Melin, M.D., Ph.D.* and Robert Dion, M.D.*

Brussels, Belgium

Sponsored by: Bruce W. Lytle, M.D., Cleveland, Ohio

Little is known about the relation between postoperative functional recovery and the ultrastructural and metabolic alterations in myocardial ischemic dysfunction. In 40 patients with anterior wall dysfunction scheduled for bypass surgery, Positron Emission Tomography (PET) with ¹³N-ammonia and ¹⁸F-deoxy glucose was performed to assess myocardial blood flow and glucose uptake, respectively. Left ventricular function and regional wall motion score were assessed by contrast ventriculography and 2-dimensional (2-D) echo-cardiography before and 6 months after surgery. Transmural biopsies were obtained from the dysfunctional myocardial area during surgery and analyzed by light and electron microscopy to quantify fibrosis (an index of irreversible ischemic damage) and ischemia-altered but viable cardiomyocytes (characterized by a loss of myofilaments and cytosolic accumulation of glycogen).

A significant correlation was found between the reduction of regional blood flow and the surface of the biopsy covered by fibrosis (r = 0.50, P<0.01). Also, increased glucose uptake at PET significantly correlated with the density of altered cardiomyocytes (r = 0.44, P<0.01) and the amount of glycogen accumulation in these cells (r = 0.42, P<0.01). Six months after surgery, the changes in regional wall motion score were analyzed in 24 patients using 2D-echocardiography: 16 patients were considered to have viable myocardium and 8 patients to have non-viable myocardium. Retrospectively, at preoperative PET, viable myocardium had displayed higher blood flow (88 ± 23 vs 60 ± 12 ml.min^-1 per 100 g, P < 0.01) and glucose uptake (50 ± 21 vs 30 ± 13 Hmol.min^-1 per 100 g, P<0.05) than non-viable myocardium. Similarly, at morphological analysis, viable myocardium had shown less tissue fibrosis (24 ± 12 vs 45 ± 21% of surface, P < 0.01) and more altered cardiomyocytes (38 ± 15 vs 24 ± 15% of surface, P < 0.05) than non-viable myocardium. At follow-up, the ejection fraction measured by contrast ventriculography was increased by 20% in patients with viable myocardium and decreased by 25% in patients with non-viable myocardium (P < 0.01).

Therefore, in patients with left ventricular ischemic dysfunction, improvement of ventricular function after surgical revascularization is associated with a higher preoperative myocardial perfusion and glucose uptake, and with less tissue fibrosis and a higher amount of viable cardiomyocytes in the dysfunctional area. Such correlation provides a morphological basis for the prognostic value of non-invasive preoperative procedures in myocardial revascularization.

*By invitation

F4. ASPARTATE/GLUTAMATE ENRICHED BLOOD DOES NOT IMPROVE MYOCARDIAL ENERGY METABOLISM DURING ISCHEMIA-REPERFUSION: A ³¹P MRS STUDY IN ISOLATED PIG HEARTS.

Hooman R. Ghomeshi, B.Sc.*, Ganghong Tian, M.D., Ph.D.*, Jian Ye, M.D.*, Jiankang Sun, M.Sc.*, Edward F. Hoffenberg, M.Sc.*, Tomas A. Salerno, M.D. and Roxanne Deslauriers, Ph.D.*

Winnipeg, Manitoba, Canada and Buffalo, New York

This study tests the effects of L-aspartate (asp) and L-glutamate (glu) with the hypothesis that depressed myocardial function and oxygen consumption (MVO₂) observed after an ischemic episode is not necessarily indicative of the lack of potential for efficient aerobic respiration, or of the need for manipulations of the perfusate as an attempt to improve such potential. ³¹P Magnetic Resonance Spectroscopy (MRS) was used to observe cellular energetics in isolated pig hearts. The hearts were perfused with blood (group A) or blood enriched with 13 mM each of asp and glu (Group B), and then subjected to 30 min normothermic, zero-flow ischemia. The hearts were then reperfused with their respective perfusates for a period of 40 min (Stage I), after which they received inotropic stimulation through a gradual increase in the perfusate (Ca⁺⁺) until the myocardial function reached a plateau and stabilized (Stage II). The data are summarized below, as percentage of pre-ischemic values. (Average pre-ischemic values for groups A and B, respectively: rate-pressure-product (RPP), 11100 and 11600 mmHg/min; -dP/dt, 1080 and 1020 mmHg/sec; MVO₂, 5.3 and 6.1 ml O₂/100g tissue/min; ATPandPCr, 100%).

*p < 0.05 with respect to stage I

The MR spectra showed no decrease in the rate of energy decline during ischemia with asp/glu enriched perfusate. Furthermore, there were no significant differences between the two groups in terms of myocardial function, oxygen consumption, or the rate or extent of high-energy phosphate recovery, after either stage I or stage II of reperfusion. There was, however, a dramatic improvement in myocardial functional parameters and a significant improvement in MVO₂ with inotropic stimulation in stage II (p < 0.05). Furthermore, despite this near-complete recovery of function following inotropic stimulation, there were no changes in levels of ATP or PCr, demonstrating the cell's ability to recruit significant aerobic (data not shown) energy production when needed. The data suggest that in an isolated pig heart subjected to 30 minutes normothermic ischemia, it is possible to inotropically recruit significant functional recovery and sufficient aerobic respiration to support it, irrespective of asp/glu enrichment.

*By invitation

F5. NEAR-ZERO COOLING: A NEW STRATEGY FOR CEREBRAL AND MYOCARDIAL PROTECTION USING A BLOOD SUBSTITUTE.

George V. Letsou, M.D.*, John H. Braxton, M.D.*, Spyros Condos, Ph.D.*, Hal Sternberg, Ph.D.*, Paul Segall, Ph.D.*, William B. Shaffer, CPP*, Hazim J. Safi, M.D. and John C. Baldwin, M.D.

Houston, Texas; New Haven, Connecticut and Berkeley, California

Profound hypothermia is being used with increasing frequency in cardiothoracic surgery. However, current techniques are limited in cooling to only 15°C for 1 hour. We examined the use of a colloid-based blood substitute to facilitate cooling animals to 1-4°C for 4 hours.

Six dogs were placed on cardiopulmonary bypass via the femoral vessels. A centrifugal pump and heat exchanger were used for systemic cooling supplemented by a cooling blanket. At 21-25°C each animal's blood was completely replaced with a colloid-based blood substitute (BioTime, Inc., Berkeley, CA). Cooling then continued to a rectal temperature of 1-4°C. Ventricular fibrillation occurred spontaneously at 17-25°C. Below 15°C, all animals were asystolic. Animals were maintained below a rectal temperature of 4°C for 4 hours. Systemic rewarming was then accomplished via the bypass circuit and a warming blanket. At 10-18°C, the colloid-based blood substitute was replaced with the animals' stored blood. All animals were weaned from cardiopulmonary bypass at 35-37°C; no animal required pressor support. All (6 of 6) were extubated and recovered normal cardiopulmonary function and behavior. Five of six animals survived 6 months to 1 year (1 died at 1 week from aspiration) with normal exercise capacity and behavior.

This blood replacement strategy allows for cooling below 4°C and results in extension of the safe period of neurologic and cardiopulmonary protection in the dog.

*By invitation

F6. ENDOTHELIAL STUNNING AND MYOCYTE RECOVERY AFTER REPERFUSION OF JEOPARDIZED MUSCLE: A ROLE OF L-ARGININE BLOOD CARDIOPLEGIA.

Asatoshi Mizuno, M.D.*, Rufus Baretti, M.D.*, Gerald D. Buckberg, M.D., Jakob Vinten-Johansen, Ph.D.*, Helen H. Young, Ph.D.* and Louis J. Ignarro, Ph.D.*

Los Angeles, California

Early sudden death sometimes follows elective or emergency operations in patients who are apparently well. Emergent placement on bypass (CPB) shows open grafts. This study tests the hypothesis that this is caused by myocyte recovery in hearts undergoing endothelial stunning (reduced nitric oxide, NO synthase activity) causing vascular hypercontraction, platelet adherence and leukocyte aggregation and subsequent muscle damage.

Thirty pigs (25 kg) underwent cardiopulmonary bypass. Six received standard glutamate/aspartate blood cardioplegia (BCP) without intermittent ischemia. Twenty-four others underwent 20 minutes of 37°C ischemia. In 6, the clamp was removed after twenty minutes, without cardioplegia infusion (injury). In 18, the clamp was left on for 30 more minutes and all received the same blood cardioplegia solution. Of these, in 6 the BCP was normal, in 6 contained L-arginine (a NO precursor) at 2 mmol/L, and in 6 contained L-NAME (an NO synthetase inhibitor), at 1 mmol/L. Measurements included contractility by LV pressure volume loops and Starling curves, NO production, endothelial dependent (acetylcholine) and independent (nitroprusside) function, and myeloperoxidase (MPO) determination for endothelial leukocyte adherence.

Complete myocyte and endothelial recovery occurred in all non-jeopardized muscle. Complete myocyte systolic and diastolic recovery occurred also in BCP and BCP + L-arginine studies (approximately 90%). Conversely, recovery was <40% in the injury group (20 minutes ischemia alone), and BCP + L-NAME group. Substantial changes occurred with NO production, as only the BCP + L-arginine group produced 85 ± 10% of normal NO. In contrast, <25% recovery was with BCP alone, and no recovery with BCP + L-NAME. Coronary endothelial dependent vasodilatation (with acetylcholine) recovered 75 ±10% with BCP + L-arginine, but <20% in all others. Conversely, independent (nitroprusside) vasodilatation was unaltered in all groups. Myeloperoxidase activity (relative to control), rose 2.3 ± 0.2 after 20 minutes of ischemia only, 1.6 ± 0.4 with BCP, and rose only 0.3 ±0.2 with BCP + L-arginine.

This discrepancy between myocyte recovery and endothelial stunning in jeopardized muscle can be offset by adding L-arginine to the blood cardio-plegic solution. This may (a) suppress production of endothelial derived contracting factor (EDCF), and (b) counteract EDCF type responses, thereby avoiding need for leukocyte filters and anti-adhesion molecules. L-arginine is a cheap, naturally occurring substance in everyone that normalizes endothelial function. This limits endothelial vasospasm and neutrophil and platelet activity, and may have major importance in subsequent cardioplegic formulations.

*By invitation

F7. ANGIOGENESIS AND COLLATERAL BLOOD VESSEL GROWTH IN CORONARY ARTERY BYPASS PATIENTS.

Roland Fasol, M.D.*, Teddy Fischlein, M.D.* and Christian Holubarsch, M.D., Ph.D.*

Bad Neustadt, Munich and Freiburg, Germany

Sponsored by: Tirone E. David, M.D., Toronto, Ontario, Canada

Background. Controlled blood vessel growth and site-directed neovessel formation in vivo by the application of angiogenic growth factors have been investigated over the past few years in experimental models. Although the application of growth factors as a possible therapeutic intervention in the management of advanced coronary artery obstructive disease may have some potential value, no clinical data are available.

Methods and Results. To assess the angiogenic potency of purified basic fibroblast growth factor (bFGF), a modified chorioallantoic membrane assay was performed, which showed significant new blood vessel growth. The growth promoting activity of purified bFGF was studied by comparing adult human endothelial cell (AHEC) control cultures, without adding growth factor to the culture medium, with AHEC cultures with 0.5, 1.0 and 5.0 ng/ml bFGF added to the culture medium. Tritiated thymidine counts proved the significant growth promoting activity of bFGF.

In a subsequent clinical study with five selected coronary artery bypass patients, 3 ml of modified fibrin glue containing 20 ng/ml bFGF was implanted between the distal internal mammary artery (IMA) pedicle and the myocardium of the anterior wall, in the area between the left anterior descending artery (LAD) and the diagonal branch. In addition, 4 ml of purified bFGF with a concentration of 20 ng/ml was applied subepicardial into the myocardium of the left ventricle in this area.

Arterial digitized computed angiography of the IMA transplanted to the LAD was performed postoperatively at the time of discharge to confirm the patency. Coronary angiography was performed after a follow-up period of one year which showed significant growth of new blood vessels in every investigated patient, indicative of coronary angiogenesis and collateral growth at the site of bFGF implantation. To quantitate the results, digital grey scale analysis of the corresponding myocardial areas in pre- and postoperative angiograms of every patient was performed. This analysis showed a significantly higher grey scale, indicative of a significantly higher density of blood vessels, at the site of bFGF implantation in the postoperative angiogram, as compared to preoperative as well as to other myocardial segments postoperatively.

Conclusions. Our data suggest that significant coronary angiogenesis and new coronary collateral blood vessel growth is possible in patients receiving angiogenic growth factor implants to the heart.

*By invitation

F8. L-ARGININE REDUCES HETEROGENEITY OF CORONARY BLOOD FLOW, REDUCES NEUTROPHIL ADHESION, AND IMPROVES RECOVERY OF VENTRICULAR FUNCTION AFTER HYPOTHERMIC ISCHEMIA

Takeshi Hiramatsu, M.D.*, Toshiharu Shin'oka, M.D.*, Marc L. Schermerhorn, M.D.*, Gregor Zund, M.D.*, Takuya Miura, M.D.*, David P. Nelson, M.D.* and John E. Mayer, Jr., M.D.

Boston, Massachusetts

Prior experiments suggest that vascular events such as the "no reflow phenomenon" are involved in ischemia (1) and reperfusion (R) injury, and histologically the injury following global ischemia is typically focal. Regional disparities (heterogeneity-Het) in post-ischemic blood flow have been demonstrated. But the relationship between vascular events and I/R myocardial injury remain unclear. We investigated the relationships between regional variations in blood flow (each LV divided into 48 pieces) using microspheres, regional neutrophil accumulation (myeloperoxidase [MPO] assayed in each LV piece), and the recovery of global ventricular function (LV max developed pressure-DP and dP/dt) in 12 isolated blood-perfused neonatal lamb hearts subjected to 2 hours of hypothermic (15°C) cardioplegic ischemia followed by reperfusion. One group received 3 mM L-arginine during R to augment endothelium mediated vasodilation and inhibit neutrophil-endothelial interactions. Controls received only cardioplegia. Endothelial function was assessed by coronary resistance response (CRR) to 10-⁷ M acetylcholine (an endothelium dependent vasodilator), and regional coronary blood flow (CBF) was measured pre-ischemia and at 5, 15, and 30 min of R. A heterogeneity index for each heart at each time point was calculated as the coefficient of variation of CBF among 48 segments (100 x SD/mean). Correlation coefficients (CC) relating CBF at 5 min of R and MPO for each of the 288 heart segments in each group were calculated (# = p<0.01; * = p<0.05 vs controls; % = % recovery of pre-ischemic values).

The inverse correlations (CC) between regional neutrophil accumulation (MPO activity) and CBF during early R suggest that microvascular occlusion by neutrophils causes regional ischemia and reduced recovery of global ventricular function. L-arginine reduced the heterogeneities in CBF between regions of the heart during early R, reduced neutrophil accumulation, and was associated with improved recovery of global LV function and endothelial function. Strategies such as L-arginine infusion which reduced neutrophil adhesion and enhance CBF during early R seem likely to be useful in limiting I/R injury.

*By invitation

9:00 a.m. SCIENTIFIC SESSIONS

Room 6, San Diego Convention Center

Moderators: G. Alexander Patterson,. M.D.

Joseph I. Miller, M.D.

15. LIVING DONOR LOBAR LUNG TRANSPLANTATION EXPERIENCE: INTERMEDIATE RESULTS.

Vaughn A. Starnes, M.D., James E. Davis, M.D.*, Mark L. Barr, M.D.*, Felicia A. Schenkel, R.N.*, Monica V. Horn, R.N.*, Winfield J. Wells, M.D.*, Jeffrey A. Hagen, M.D.*, Robbin G. Cohen, M.D.* and The University of Southern California Lung Transplant Group

Los Angeles, California

Discussant: Charles B. Huddleston, M.D.

Living donor lobar lung transplantation offers an alternative form of treatment for patients with end-stage lung disease who might otherwise die while waiting for traditional cadaveric lung transplantation. Thirty-four adult (25) and pediatric (9) patients (14 females, 20 males) have undergone living lobar transplantation since January 1993. The mean age was 22.5 ± 1.1 yrs (range 9.3 to 36 yrs). Thirty-one patients had cystic fibrosis. Other indications were pulmonary hypertension, pulmonary fibrosis and obliterative bronchiolitis. Patients were selected for transplantation based on increasing frequency or severity of infections, as well as developing antibiotic resistance or deteriorating pulmonary function tests. The mean weight was 43.7 ± 1.7 kg. The mean pCO₂ and pO₂ were 65 ± 5 and 70 ± 5 mmHg, respectively. Preoperatively, 23 patients required supplemental oxygen or mechanical ventilation. Twenty-five procedures were done emergently or urgently. The mean lobe ischemic time was 65 ± 6 minutes. Immunosuppression consisted of cyclosporine, azathioprine, and corticosteroids. Gancyclovir was used in 28 cases where either the recipient or donor was cytomegalovirus (CMV) positive. The mean survival time was 20.2 ± 2.1 months with a 1-year survival of 66.1 ± 9.1%. There were 7 in hospital deaths and 4 late deaths. Rejection occurred in 16 patients (0.07 rejections/patient month) and were treated with steroid pulse therapy. Twenty-three patients developed post-operative infections which were most often Pseudomonal pneumonias but included 7 patients with fungal infections and 7 with CMV infections. Postoperatively, pulmonary function tests improved significantly (p<0.001). FEV₁ and FVC increased from 23 ± 2% to 68 ± 4% and 37 ± 2% to 67 ± 4% of predicted, respectively. The most significant improvement occurred in the FEF25.75 measurements (8.5 ± 2.2% to 68 ± 6%). Normal cardiac and pulmonary hemodynamics were present postoperatively. Fourteen patients were catheterized. Mean right atrial pressure, pulmonary wedge pressure, pulmonary artery systolic and diastolic pressures were 4 ± 1 mmHg, 10.4 ± 1.6 mmHg, 32.1 ± 2.7 mmHg, and 13.4 ± 1.7 mmHg, respectively. The mean NYHA class improved from 3.1 ±0.1 to 1.1 ±0.1 (p<0.001). These intermediate results demonstrate that living donor lobar lung transplantation is in effective therapy for critically ill patients in need of urgent transplantation.

*By invitation

16. SINGLE OR BILATERAL LUNG TRANSPLANTATION FOR EMPHYSEMA?

Sudhir Sundaresan, M.D.*, Yuji Shiraishi, M.D.*, Jenny Manley, R.N.*, Dottie Bigger, R.N.*, Mary Pohl, R.N.*, Elbert P. Trulock, M.D.*, Joel D. Cooper, M.D. and G. Alexander Patterson, M.D.

St. Louis, Missouri

Discussant: J. Kent Trinkle, M.D.

The optimal lung transplant procedure for Chronic Obstructive Pulmonary Disease (emphysema and Alpha-1 Antitrypsin Deficiency) continues to be a matter of controversy. To date, no long-term follow-up study with comparison of late results is available. Although most programs favor single lung transplantation (SLT), we have limited SLT mainly to older patients and those with small body habitus. We analyzed our experience with 118 consecutive lung transplants for emphysema between 1989-1994 (50 SLT and 68 bilateral lung transplants [BLT]) to try to determine which might be the superior option. The SLT group was older and had a higher proportion of females (68% vs. 43% in BLT group, p < 0.01). However, pulmonary function (FEV₁), arterial oxygen tension (PaO2), and exercise tolerance (distance covered in a 6-minute walk test) were similar. Post-transplantation, 90-day mortality (SLT 10% vs BLT 8%; p = 0.74), and duration of mechanical ventilation, intensive care unit stay, and hospitalization were similar. A significant improvement in FEV_1; PaCO2, PaO2, and exercise tolerance was noted within 3 months post-transplantation in both groups, and was sustained. However, the improvements in FEV₁, PaO2, and exercise tolerance were consistently and significantly better in the BLT group at and beyond 6 months (data shown in table).

Obliterative bronchiolitis (OB) was equally prevalent in both groups (BLT 37% vs SLT 38%; p = 0.99). Survival was similar in both groups until 3 years (BLT 83% vs SLT 81%), but subsequently showed a tendency to better survival in the BLT group (5 year actuarial survival: BLT 65% vs SLT 42%). These data demonstrate that: (1) both SLT and BLT are satisfactory options in emphysema and produce durable results; (2) BLT is not associated with
increased operative mortality or morbidity, and produces significantly better improvement in spirometry, oxygenation, exercise tolerance, and possibly late survival than SLT; and (3) the observed benefit following BLT on late survival and function may be attributed to the presence of more pulmonary reserve after the onset of OB.

*By invitation

17. LONG-TERM RESULTS OF LUNG METASTASECTOMY: REPORT ON 5,206 CASES FROM THE INTERNATIONAL REGISTRY OF LUNG METASTASES.

Ugo Pastorino, M.D.*, Robert Ginsburg, M.D., Patricia McCormack, M.D., Joe Putnam, M.D., Harvey Pass, M.D., Michael Johnston, M.D.* and Peter Goldstraw, M.D.*

London, United Kingdom; New York, New York; Houston, Texas; Bethesda, Maryland and Toronto, Ontario, Canada

Discussant: Valerie W. Rusch, M.D.

An International Registry of Lung Metastases was established in 1991 to assess the long-term results of pulmonary metastasectomy. Data has been collected from 18 major departments of thoracic surgery in Europe (13), United States (4) and Canada.

The Registry has accrued a total of 5207 cases of lung metastasectomy, of which 4572 (88%) underwent complete surgical resection. Main patient features were: 2932 males and mean age 44 (range 2-93). In 2260 cases the pulmonary metastases were from an epithelial tumour, 2203 from sarcomas, 363 had germ cell deposits and 328 melanomas. 1603 cases experienced a disease-free interval (DPI) of 0-11 months (mos), 1857 cases had DPI 12-35 mos and 1620 cases had DPI 36+ mos. There were 3687 thoracotomies and 1415 sternotomies. Single proven metastases accounted for 2383 cases, and 2726 had multiple lesions. Mean follow-up of the whole series was 46 mos. Analysis of the data, including Kaplan-Meier estimates of the survival curves, related risks of death (RR), proportional hazard and multivariate Cox model was performed by an independent centre in Brussels.

The survival after complete metastasectomy was 36% at 5 years, 26% at 10 years and 22% at 15 years, with a median survival of 35 mos; the corresponding values for incomplete resection were 13% at 5 years and 7% at 10 years, with a median of 15 mos. Among complete resections, the 5-year and median survival was 33% and 29 mos for patients with DPI of 0-11 mos, 31% and 30 mos for DPI of 12-35 mos, 45% and 49 mos for DPI of 36+ mos. The 5-year and median survival for complete resection of single lesion cases was 43% and 43 and mos, 34% and 31 mos for 2-3 lesions, 27% and 27 mos for 4+ lesions.

At multivariate analysis, the primary tumor type, DPI and number of metastases emerged as significant prognostic factors. In particular, germ cell and Wilms' tumors showed the best prognosis (RR=0.4) and melanoma the worst prognosis (RR=2.1); DPI of 36+ mos had a better prognosis (RR=0.6), as well as single metastases (RR=0.7).

On this basis, further analyses are in progress with the aim of defining a system of classification in three to four prognostic groups, valid for the various primary tumor types.

10:00 a.m. INTERMISSION - VISIT EXHIBITS

10:45 a.m. SCIENTIFIC SESSIONS

Room 6, San Diego Convention Center

Moderators: Mortimer J. Buckley. M.D.

Fred A. Crawford, Jr., M.D.

18. SUPERIOR VENA CAVA TO PULMONARY ARTERY ANASTOMOSIS: AN ADJUNCT TO BIVENTRICULAR REPAIR.

Glen S. VanArsdell, M.D.*, William G. Williams, M.D., Catherine M. Maser, R.N.*, Kim Streitenberger, R.N.*, Ivan M. Rebeyka, M.D.*, John G. Coles, M.D. and Robert M. Freedom, M.D.*

Toronto, Ontario, Canada

Discussant: Frank L. Hanley, M.D.

A morphologically small or poorly functioning right ventricle (RV) can increase the risk of biventricular repair. It has been proposed that a superior vena cava to pulmonary artery (SVC-PA) anastomosis can unload an inadequate RV thereby enhancing the success of biventricular repair. From May 1981 to September 1995, 41 patients received a SVC-PA anastomosis in association with biventricular repair. Indications were: group A (19 pts), small physiologic right ventricle defined as tricuspid annulus z value of < -2 ([mean -3.7 ±1.4] [14/19]) or predicted RV volume of 69-87% ([mean 77.8 ± 8] [5/19]); group B ([11 pts] [9/11 Ebsteinis]), chronic RV dysfunction; group C (4 pts), facilitation of repair (e.g. LSVC and AVSD); group D (4 pts), postoperative RV dysfunction. Three patients were excluded from analysis because the SVC-PA anastomosis was unrelated to biventricular repair. Age ranged from 5 months to 51 years (median 3.5 years). Diagnostic subsets were heterogenous: Ebsteinis anomaly (9), d-TGA(4), L-TGA(4), VSD (3), TOP (3), PS (3), pulmonary atresia with IVS (3), ASD (2), DORV (2), AVSD (2), DILV with TGA (1), AVSD with TOP (1), and left atrial isomerism complex (!)* Twenty-five patients received a right bidirectional cavopulmonary anastomosis (BCPA), 6 a side-to-side cavopulmonary anastomosis, 4 a left BCPA, 2 a classic Glenn shunt, and 1 a bilateral BCPA anastomosis. Operative mortality was: group A 2/19 (10.5%), group B 1/11 (9%), group C 0/4 (0%) and group D 3/4 (75%). Follow-up is complete in 37/38 patients (97%), ranging from 1 to 174 months (mean 45.6 ± 37.3) with one late mortality (1/31, 3%). NYHA class is I (22) or II (8). There is no clinical evidence of pulmonary AV fistulae or protein losing enteropathy. In the setting of an inadequate RV, this series demonstrates an acceptable intermediate term outcome for SVC-PA anastomosis and biventricular repair. In a subset, SVC-PA anastomosis safely facilitates repair. Results for this procedure when used as a salvage operation for severe postoperative RV dysfunction have not been satisfactory.

*By invitation

19. IS A HIGH RISK BIVENTRICULAR REPAIR ALWAYS PREFERABLE TO CONVERSION TO A SINGLE VENTRICLE REPAIR?

Ralph E. Delius, M.D.*, Marc A. Radermecker, M.D.*, tMarc R. de Leval, M.D., Martin J. Elliott, M.D.* and Jaroslav Stark M.D.

London, United Kingdom

Discussant: Vaughn A. Starnes, M.D.

AIM: A commonly held concept is that in the presence of two ventricles, a two ventricle repair is always preferable over a single ventricle repair. Whether this policy should be adhered to for high-risk biventricular repair remains controversial. The aim of this report is to examine the short and intermediate term outcome of a complex two ventricle repair with a single ventricle repair in patients with two functional ventricles.

PATIENT POPULATION: Since 1986, 34 patients with atrioventricular (AV) concordance or discordance, ventriculoarterial (VA) discordance, ventricular septal defect (VSD), and pulmonary stenosis or atresia (PS/PA) have undergone VSD closure and placement of a valved conduit between the pulmonary ventricle and pulmonary artery (Group I). Another group of 16 patients (Group II) with the same diagnoses have undergone a single ventricle repair consisting of a total cavopulmonary connection (TCPC) because of either a straddling AV valve (11 patients) or an uncommitted VSD (5 patients). These patients would have otherwise been candidates for a two ventricle repair.

RESULTS: The mean length of follow-up in the Group I was 3.9 years (range 1-10 years) and 3.0 years in Group II (range 0.5-9 years). There were 12 late reoperations in 11 patients in Group I and no late reoperations in Group II. Freedom from reoperation at 5 years was 60.5% in Group I and 100% in Group II (p < 0.03). There were 5 early and 3 late deaths in Group I. There was one early death in Group II and no late deaths. The actuarial estimate of survival at 5 years was 68.0% in Group I and 93.8% in Group II (p < 0.05). In Group I there were 2 patients (6.5%) that were Class III or Class IV; the remainder were Class I or II. All patients in Group II were NYHA Class I or II.

CONCLUSION: There appeared to be greater short and intermediate term morbidity and mortality in patients who underwent a conventional two ventricle repair compared to a similar group of patients who underwent a TCPC, even though the patients managed with a two ventricle repair were ideal candidates for this approach in contrast to the patients managed with a TCPC, who would have been at high risk for a two ventricle repair because of a straddling AV valve or an uncommitted VSD. It is unknown if the funtional long-term results of a TCPC in patients with 2 ventricles are as good as those obtained with a two ventricle approach. However, these findings suggest that there may be situations in which the short and intermediate term risks of a complex two ventricle repair may outweigh the long-term disadvantages of a single ventricle approach.

11:25 a.m. HONORED GUEST SPEAKER

Tolerance to Allogeneic and Xenogeneic Transplants

David H. Sachs, M.D., Boston, Massachusetts

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

12:10 p.m. CARDIOTHORACIC RESIDENTS' LUNCHEON

†1973-74 Graham Fellow

*By invitation