TUESDAY AFTERNOON, APRIL 26, 1994

1:45 p.m. SIMULTANEOUS SCIENTIFIC SESSION A - ADULT CARDIAC SURGERY

East Ballroom

Moderators: John L. Ochsner, M.D.

Tirone E. David, M.D.

18. LONG TERM RESULTS OF VALVE REPLACEMENT WITH THE ST. JUDE MEDICAL PROSTHESIS

Eugene M. Baudet, M.D.*, Vincent Fuel, M.D.*, Francois Roques, M.D.*, Frederic Clerc, M.D.*, Xavier Roques, M.D.* and Nadine Laborde, M.D.*

Bordeaux, France

Sponsored by: D. Craig Miller, M.D., Stanford, California

Since June 12, 1978, the St. Jude Medical (SJM) valve has been routinely used as our mechanical prosthesis of choice; through September, 1993, 2, 714 SJM valves were implanted.

To assess results with truly long follow-up, we reviewed the first 1, 112 patients (pts) undergoing 1, 244 valve replacements with the SJM prosthesis before June 12, 1987: Aortic (AYR) = 773 pts (69%), Mitral (MVR) = 207 pts (19%) or Mitral and Aortic position (DVR) = 132 pts (12%). The mean patient age was 55.9 years (range 9 months to 82 yrs) and 690 pts (62%) were males. There were 42 hospital deaths (3.8%). Follow-up was 98% complete (8, 988 pt-yrs), with a mean of 9.75 yrs, range 6-15 yrs. There were 213 late deaths: 91 or 43% (60 AYR, 19 MVR, 12 DVR) were considered valve-related. Eleven (12%) were due to prosthetic valve endocarditis (PVE), 27 (30%) to sudden death (SD), 9 (10%) to valve thrombosis (VT), 19 (21%) to thromboembolism (TE), 22 (24%) to anticoagulant-related hemorrhage (ACRH), and 3 (3%) to paravalvular leak (PVL). Actuarial survival, at 14 yrs, including hospital mortality, was 69 ± 7% for AVR, 68 ± 11% for MVR and 59 ± 16% for DVR. Linearized rates of late valve-related events included: TE (1.09% pt-yr), ACRH (0.94% pt-yr), PVE (0.32% pt-yr), VT (0.33% pt-yr), PVL (0.34% pt-yr). Actuarial freedom, at 14 yrs, from TE was 89 ± 3%, ACRH (83 ± 8%), VT (97 ± 1%), reoperation (95 ± 3%). Long term actuarial freedom from all valve-related deaths was 84 ± 6%, and 61 ± 8% from all valve-related morbidity and mortality. At follow-up, 93% of the survivors were in NYHA class I or II.

With one of the longest follow-up experiences reported for this prosthesis, the St. Jude valve has proven, because of its low thrombogenicity, low incidence of valve-related events, and low valve-related mortality, to be one of the best performing mechanical prosthesis currently available. Nevertheless, the late valve-related complications and deaths illustrate how our quest for a "perfect" prosthesis still remains unfulfilled.

*By invitation

19. THE CARPENTIER-EDWARDS PERICARDIAL AORTIC VALVE; TEN YEAR RESULTS

Delos M. Cosgrove, M.D., Bruce W. Lytle, M.D.*, Paul C. Taylor, M.D.*, Margarita T. Camacho, M.D.*, Robert W. Stewart, M.D.*, Patrick M. McCarthy, M.D.* and Floyd D. Loop, M.D.

Cleveland, Ohio

The utilization of bioprostheses in the aortic position has been limited by the restrictive hemodynamics and suboptimal durability of porcine valves. Pericardial valves solved the hemodynamic problems but most pericardial valve designs demonstrate an unacceptably high rate of structural valve deterioration. To evaluate the function of the Carpentier-Edwards pericardial valve in aortic position the results of 310 aortic valve replacements performed between 1982 and 1985 were analyzed. Mean age was 64.2 ± 10.8 years (range 22-95); 190 (61.3%) were males. Isolated aortic valve replacement was performed in 272 patients (87.7%). There were 18 hospital deaths (5.8%) and none were valve related.

Follow up of the 292 survivors was 100% complete at a mean of 8.8 ± 0.8 years; 2, 209 patient years of follow up were available for analyses. There were 119 (40.7%) late deaths. Actuarial survival at 5 and 10 years were 82.5% and 55.2% respectively.

The 153 patients ≥65 had an extremely low incidence of structural valve deterioration with only 4 explants and 95.1% actuarial freedom from explant at 10 years and a linearized rate of 0.4 ± 2.3% per patient year compared to 83.5% and 0.9 ± 3.2 for patients <65.

Fourteen valves were explanted for structural deterioration. Of these 13 (93%) had leaflet calcification causing stenosis and one had a wear-related leaflet tear.

We conclude that the Carpentier-Edwards pericardial valve 1) has a low incidence of valve-related complications; 2) structural deterioration is infrequent and results from leaflet calcification. 3) The low incidence of structural deterioration in patients ≥65 makes this valve an increasingly appropriate option in this group.

*By invitation

20. MID-TERM RESULTS OF AORTIC VALVE REPLACEMENT WITH STENTLESS PORCINE AORTIC VALVE

Christopher M. Feindel, M.D.*, Tirone E. David, M.D., Joanne Bos, R.N.*, Zhao Sun, M.A.*, Susan Armstrong, MSc.* and Hugh E. Scully, M.D.

Toronto, Ontario, Canada

A stentless porcine valve (SPV) has been used for aortic valve replacement (AYR) in 114 pts since 1987. There were 83 men and 31 women whose mean age was 61 ± 10 years. Eleven pts were in chronic atrial fibrillation; 85 had aortic stenosis; 77 were in NYHA functional class 3 or 4, and 39 had coronary artery disease. The SPV was secured in the subcoronary position using the same technique as for free-hand AY homograft. The mean SPV size was 26.4 mm (range 19 to 29 mm). There was only one operative death due to myocardial infarction but 6 pts had serious postoperative complications. Pts have been followed from 2 to 72 months, mean of 26. There have been 2 late deaths, neither one was valve-related. The actuarial survival at 5 years was 92% ± 3%. There have been only 3 valve-related complications: one infective endocarditis and two TIAs. Doppler echocardiographic studies have been performed annually; 89 pts have no AY insufficiency and 20 have mild. The mean AY area is 1.78 cm.sq. (range 1.1 to 3.0) and it has remained unchanged up to six years.

The SPV pts were individually matched with 376 pts who had AYR with Hancock II (HAN) bioprosthesis. The following variables were used for case-matching: age (±5 years), valve lesion, NYHA functional class, left ventricular ejection fraction (± 5%), and coronary artery disease (# of vessels diseased). Of 114 SPV pts, 101 were case-matched with 101 HAN pts. There was no operative mortality in matched pts. The actuarial survival at 5 years was 93% ± 4% for SPV pts and 86% ± 5% for HAN pts (p = ns). Proportional hazard analysis revealed that valve-related complications were three times more common in HAN pts than in SPV pts.

These data suggest that AYR with SVP is associated with a lower rate of valve-related complications than AYR with HAN. For this reason, AYR with SPV may enhance pts survival.

2:45 p.m. INTERMISSION - VISIT EXHIBITS

*By invitation

3:15 p.m. SIMULTANEOUS SCIENTIFIC SESSION A - ADULT CARDIAC SURGERY

East Ballroom

Moderators: John L. Ochsner, M.D.

Tirone E. David, M.D.

21. RETROGRADE CARDIOPLEGIA DOES NOT PERFUSE THE RIGHT VENTRICLE

Bradley S. Allen, M.D.*, Renee S. Hartz, M.D., Jacqueline Wiewall, B.S.*, Hanafy Hanafy, M.D.*, Solomon Aronson, M.D.*, Laurence Segil, M.D.*, David Mayer, M.D.*, Steven Feinstein, M.D.*, Kirk Boiling, M.D, M.P.H.* and Jongwok Ham, B.S.*

Chicago, Illinois

Surgeons often rely primarily on retrograde cardioplegia for myocardial protection, since it provides adequate left ventricular (LV) perfusion, even in the presence of coronary artery disease. Clinically, however, adequate right ventricular (RV) perfusion by retrograde delivery, has not been demonstrated. Utilizing intraoperative TE echo, we examined retrograde delivery of cardioplegic solutions by contrast echocardiography which directly assesses myocardial perfusion. Fifteen patients (7CABG, SValves) had 4 cc's of sonicated Isovue® injected retrograde via a coronary sinus catheter. Quantitive assessment of myocardial perfusion was performed by visual inspection and background subtracted videodensitometric analysis. Prior to removing the aortic cross-clamp, myocardial oxygen extraction was calculated by first delivering 2 mins of warm blood cardioplegia retrograde, and then taking samples from the cardioplegic line and aortic root. This determined the oxygen extraction ratio across the myocardium at the end of retrograde delivery. Warm blood cardioplegia was next given antegrade, and 15 sec's later samples taken from the cardioplegic line and a right ventricular (acute marginal) vein, to determine the oxygen extraction ratio across the RV. As assessed by contrast echo, retrograde infusion resulted in almost four times greater perfusion to the left ventricular free wall and septum, compared to the RV free wall (See graph).

Oxygen extraction across the myocardium supplied by retrograde infusion was low after two minutes. Conversely, when antegrade cardioplegia was started, R.V. oxygen extraction rose 4 fold (42 ± 5% vs 11 ± 1%, p<0.05) demonstrating that retrograde cardioplegia had not adequately perfused the RV myocardium.

Conclusions: 1. Retrograde cardioplegia provides poor R.V. myocardial perfusion as assessed by contrast echocardiography. (2) This poor perfusion is inadequate to meet myocardial demands as demonstrated by the high R.V. oxygen extraction after a prolonged retrograde infusion. (3) Therefore, surgeons must not rely solely on retrograde cardioplegia for RV myocardial protection. This concept is especially important if continuous warm blood cardioplegia is used, as myocardial requirements are then higher.

mean ± S.E.

*By invitation

22. THE CAPILLARY DISTRIBUTION OF RETROGRADE BLOOD CARDIOPLEGIA IN EXPLANTED HUMAN HEARTS

Abbas Ardehali, M.D.*, Hillel Laks, M.D., Richard N. Gates, M.D.*, Davis C. Drinkwater, M.D.*, Thomas J. Sorensen, B.S.* and Paul Chang, B.S.*

Los Angeles, California

Warm retrograde blood cardioplegia (RBCP) for myocardial protection is frequently used despite several experimental animal studies questioning the adequacy of capillary flow to the right ventricle (RV) and septum. The capillary distribution of RBCP in the human heart is unknown. Hearts from 8 transplant recipients with the diagnosis of idiopathis/viral cardiomyopathy were arrested in situ with cold blood cardioplegia and excised with the coronary sinus intact. Within 20 minutes of explantation, colored microspheres (15 ± 5 µm) mixed in 37°C blood cardioplegia were administered through the coronary sinus at a pressure of 30-40 mm Hg for 2 minutes. Twelve transmural myocardial samples were taken horizontally at the level of midventricle and apex to determine regional capillary flow rates.

Results:

At 37°C, approximately 18cc/100gm/min of capillary blood cardioplegia flow is needed to meet the metabolic requirements of the arrested human heart (Assuming a hemoglobin of 8 gm/dL, 100% O₂ saturation, and 60% O₂ extraction). Conclusions: In human hearts, RBCP delivered at a pressure of 30-40 mm Hg provides adequate capillary flow to the LV, septum, apex, and posterior wall of RV. However, the capillary flow to the anterior and lateral wall of RV is marginal. These findings suggest that the RV of the human heart may not be adequately protected by warm RBCP.

*By invitation

23. LONG-TERM FOLLOWUP OF 7551 CORONARY ARTERY BYPASS GRAFTS: FACTORS INFLUENCING PATENCY

William H. Coltharp, M.D.*, Michael D. Decker, M.D.*, William S. Stoney, M.D., William C. Alford, Jr., M.D., George R. Burrus, M.D.*, David M. Glassford, Jr., M.D.*, John W. Lea, IV, M.D.*, Michael R. Petracek, M.D.*, Todd A. Shuman, M.D.* and Thomas D. Starkey, M.D.*

Nashville, Tennessee

In order to assess the influence of patient characteristics (smoking history, diabetes mellitus, cholesterol history, age, sex, weight, and body surface area) and surgical technique (choice of conduit, choice of graft type, and vessel bypassed) on the long-term patency of coronary artery bypass (CAB) grafts, we evaluated surgical and catheterization data for 2317 patients who had at least one cardiac catheterization subsequent to coronary artery bypass. Duration of followup from CAB to last catheterization ranged from 7 days to 20.2 years (mean, 5.3 years) for the 7551 evaluable grafts. Graft patency was evaluated by Kaplan-Meier product-limit survivor analysis; cofactors were evaluated by Cox proportional hazards regression.

The overall graft patency rates were: 1 year. 96%; 2 years, 94%; 5 years 83%; 10 years 51%; 15 years 27%; and 20 years, 11%. These are conservative estimates, as they include only those patients who presented for repeat catheterization following CAB.

Patient characteristics that were significantly associated with duration of graft patency were sex, weight, and ever having smoked. Body surface area showed a significant association only when weight was not included in the model; history of diabetes was no longer significant once weight and sex were in the model.

Technical factors correlating significantly with graft survival were choice of conduit, graft type, and vessel bypassed. Graft patency rates were best for the anterior descending (AD) artery; intermediate for the acute marginal, circumflex, diagonal, obtuse marginal, posterior descending, posterior ventricular, right main and intermediate arteries; and least satisfactory for the left main and septal perforator arteries (patency rates at 10 years, 65%, 45% and 33% respectively; see Table). Internal mammary artery (IMA) grafts were associated with significantly better patency rates than were greater saphenous vein (GSV) grafts, an effect only partially explained by their more frequent use for AD grafts. Ten year survival rates were 53% for single grafts, 49% for sequential grafts, 47% for natural "Y" grafts, and 26% for constructed "Y" grafts.

Conclusions. Technical choices of conduit, graft type, and vessel to be grafted significantly influence CAB graft patency rates. The internal mammary artery is the conduit of choice. When grafting multiple vessels with one graft, the sequential technique appears to offer better graft survival than do "Y" grafts.

*By invitation

24. CORONARY ARTERY BYPASS GRAFTING WITH THE INFERIOR EPIGASTRIC ARTERY (IEA): MIDTERM CLINICAL AND ANGIOGRAPHIC RESULTS

Michel Buche, M.D.*, Jean-Claude Schoevaerdts, M.D.*, Erwin Schroeder, M.D.*, Olivier Gurne, M.D.*, Yves Louagie, M.D.* and Baudouin Marchandise, M.D.*

Yvoir, Belgium

The IEA has been recently used for CABG operations. Histological similarities with the Internal Mammary Artery (IMA) suggest that its patency rate will be comparable to that of free IMA grafts; however, no study has yet validated this assumption.

Material: Between December 1988 and September 1993, 157 patients (141 males, 16 female, mean age: 60.2 years, range 37 to 78 years) underwent a complete myocardial revascularization using 157 IEA, 285 IMA (281 in situ, 4 free grafts). A total of 543 arterial anastomoses (mean 3.4, range 2 to 5 per patient) were constructed, among them 167 with the IEA which was anastomosed to 2 LAD, 5 diagonal, 34 circumflex and 126 RC arteries. The indication for use of the IEA was reoperation in 14 patients, varicose or stripped vein or peripheral arteritis in 42 and a favorable anatomy in 101 selected patients. An early recatheterization was obtained before discharge (mean 11 days postoperatively) in 135 patients, among whom 77 underwent a later angiographic study 6 to 43 months after surgery. A complete follow-up is available for all the patients.

Results: The follow up averages: 26 months (range: 1 to 58 months). Four patients died early and there were 3 perioperative non fatal myocardial infarction (MI). Seven patients required early reoperation for thoracic bleeding (1) or drainage of an abdominal parietal collection (6). There were 4 late deaths (2 sudden deaths, 2 non cardiac causes) and one non fatal MI. Angina recurred in 9 patients of whom 1 required reoperation and 3 underwent successful PTC A of a native coronary artery.

132/135 of the IEA were patent at early control of those 12 showed segmental irregularities or stenotic anastomosis. 44/48 of the IEA restudied within the first postoperative year (mean 8.5 months) were patent however 8 showed a diffuse narrowing. 28/29 of the IEA controlled between 13 and 43 months (mean 25 months) are open and among those 25/29 are widely patent, perfectly matching with the receiving coronary artery. It has to be emphasized that most of the occluded or narrowed IEA were grafted onto coronary arteries with mild stenosis at restudy.

Conclusion: The early attrition rate of the IEA, as for any free arterial graft is probably the result of both the loss of a true pedicle and the need for constructing an additional proximal anastomosis. The improved patency rate beyond one year could suggest a better durability in the future.

4:45 p.m. EXECUTIVE SESSION (Members Only)

East Ballroom

*By invitation

TUESDAY AFTERNOON, APRIL 26, 1994

1:45 p.m. SIMULTANEOUS SCIENTIFIC SESSION B - GENERAL THORACIC SURGERY

Trianon Ballroom

Moderators: John R. Benfield, M.D.

J. Kent Trinkle, M.D.

25. BILATERAL PNEUMECTOMY (VOLUME REDUCTION) FOR CHRONIC OBSTRUCTIVE PULMONARY DISEASE

Joel D. Cooper, M.D., Elbert P. Trulock, M.D.*, G. Alexander Patterson, M.D., Anastasios Triantafillou, M.D.*, R. Sudhir Sundaresan, M.D.*, Carolyn M. Dresler, M.D.* and Charles L. Roper, M.D.

St. Louis, Missouri

Surgical options for the treatment of emphysema have included bullectomy for patients with compression of normal underlying lung, and lung transplant for patients with severe disability and limited life expectancy. Based upon observations and experience published 34 years ago by Brantigan, we have undertaken bilateral lung volume reduction surgery in 8 patients with severe emphysema to reduce total thoracic volume and improve chest wall and lung mechanics. All patients had severe functional disability, a distended thorax, and no normal, compressed lung as judged by CT scan. Age range was 39 to 76 years (mean 56). The operation consists of median sternotomy and non-anatomic resection of multiple pieces of lung from either side using a linear stapling device. Portions of lung excised include both bullous and non-bullous areas. All patients have been extubated at the end of the procedure and there has been no early or late mortality. Two patients required re-exploration, one for bleeding and one for persistent air space. Hospital stay was 7 to 48 days (median 15) with persistent air leak the major source of morbidity. Recent use of a buttressed staple line has significantly reduced this complication. There were no sternal wound infections. Followup ranges from 3 to 10 months (mean 6 months).

RESULTS: (Mean values)

Supplemental oxygen, required pre-operatively by 4 patients either at rest or with exertion, is not required by any patient. All patients have experienced a dramatic improvement in their quality of life. The early improvement observed in the FEV₁, has been sustained or further improved in all patients. Our experience to date, though very preliminary, suggests that volume reduction surgery may be a useful therapeutic modality in carefully selected patients, including some for whom this procedure may be a "bridge" to future lung transplantation.

*By invitation

26. SS-LOBECTOMY: A SAFE TECHNIQUE FOR VATS

Ralph J. Lewis, M.D.

New Brunswick, New Jersey

Currently, VATS techniques are being borrowed from the open conventional thoracotomy, however, these same techniques have made VATS lobectomy difficult, burdensome and even dangerous. SS-Lobectomy (Simultaneous Stapling of all hilar structures in their natural anatomical configuration), has been performed successfully in 16 patients. Every attempted SS-Lobectomy is included. There were 14 malignancies, 1 giant benign pulmonary cyst and 1 large necrotizing granuloma. Three RUL, 6 RLL, 4 LUL, 2 LLL and 1 RML were resected uneventfully. Nine adenocarcinoma, 2 large cell carcinomas and 3 squamous cell carcinomas ranging in size from 2.5 to 5 cms were removed. Lung fissures, hilum and mediastinum were explored for lymph nodes in each patient. Median operative time was 110 minutes. Average blood loss was less than 100 ccs. Median hospitalization was 6 days, however, eight patients were discharged between 3 and 5 days. Three patients had air leaks averaging 14 days and one patient had mild subcutaneous emphysema for 5 days. There was no surgical mortality. Median follow-up is 15 months (range 8 to 20 months). SS-Lobectomy is not meant to replace the conventional lobectomy by open thoracotomy. Indications are cardiac or renal problems, contralateral chest wall paralysis, neurogenic deficiencies, adamant refusal of open lobectomy (psychological aberrations, pain from a previous thoracotomy). Contraindications include absent fissures, enlarged matted invasive nodes, fibrotic hilum, central or bulky lesions, calcific bronchi, chest wall invasion or lesions crossing a fissure. Precedent for this technique will be discussed.

Possibly, when used with discretion in certain carefully selected patients, in whom an open lobectomy would be contraindicated, SS-Lobectomy might eventually prove to be another available option. Time and further experience will be necessary to determine its true merits.

*By invitation

27. SUBXIPHOID PERICARDIAL WINDOW FOR PERICARDIAL TAMPONADE: SAFE, COST EFFECTIVE AND DURABLE

Darroch W.O. Moores, M.D.*, Keith B. Allen, M.D.*, David J. Gillman, P.A.*, William H. Warren, M.D. Stanley W. Dziuban, M.D.*, Riivo Ilves, M.D.* and L. Penfield Faber, M.D.

Albany, New York and Chicago, Illinois

Due to recent reports and enthusiasm for VATS pericardectomy we reviewed our experience with subxiphoid pericardial window. From 8/15/88 to 6/7/93 155 patients underwent subxiphoid pericardial window for pericardial effusion associated with pericardial tamponade. There were 85 females (55%) and 70 males ranging in age from 5 weeks to 88 years. The procedure was carried out under general anesthesia in 113 patients (72%), local/sedation in 42 patients. Underlying malignancy was present in 82 patients, 73 patients had benign disease. Follow-up is complete in all patients. The overall 30 day mortality was 20%. The 30 day mortality in patients with malignancy was 32.9% (27/82) versus 5.4% (4/73) for patients with benign disease. None of the post-operative deaths were attributed to the surgical procedure. Recurrent pericardial effusion requiring further surgical intervention occurred in four patients (2.5%), two with malignancy (2.4%) and two with benign disease (2.7%). Median survival in patients with benign disease was 482 days versus 83 days in patients with malignancy. (P=<.01) Median survival in patients with malignancy who had proven malignant pericardial effusion was 54 days compared to 95 days for patients with malignancy who did not have tumor in the pericardium. (P=<.01) Conclusion: 1. Subxiphoid pericardial window is the procedure of choice for patients with pericardial effusion and pericardial tamponade. It can be done under local anesthesia and does not require single lung anesthesia for collapse of the lung. These are important considerations in critically ill patients with pericardial tamponade. 2 Transthoracic pericardial window by open or video assisted technique offers no benefit over the subxiphoid approach. 3. Following pericardial drainage, survival is significantly shorter in patients with malignant pericardial effusion compared to patients with malignancy who do not have tumor in the pericardium.

2:45 p.m. INTERMISSION - VISIT EXHIBITS

*By invitation

3:15 p.m. SIMULTANEOUS SCIENTIFIC SESSION B - GENERAL THORACIC SURGERY

Trianon Ballroom

Moderators: John R. Benfield, M.D.

J. Kent Trinkle, M.D.

28. CRITICAL ISSUES IN PEDIATRIC LUNG TRANSPLANTATION

John M. Armitage, M.D., Geoffrey Kurland, M.D.*, Marian Michaels, M.D.*, Lynne Cipriani, R.N.*, F. Jay Fricker, M.D.* and Hartley P. Griffith, M.D.

Pittsburgh, Pennsylvania

Lung transplantation in children is a feasible, safe and effective form of therapy for end stage lung disease. Forty children (age 1-18 years, 27 female and 13 male) have undergone heart-lung(21), double lung(17) and single lung(2) transplant procedures at our center from 1985 through October 1993. The indications for transplantation have been diverse, primary pulmonary hypertension (10), cystic fibrosis (CF) (11), congenital heart disease (CHD) (10), arteriovenous malformation (3), emphysema (1), graft versus host disease (1), rheumatoid lung (1), cardiomyopathy (1), desquamative interstitial pneumonitis (1) and Proteus syndrome (1). The actuarial 1 year survival was 75% (mean follow-up 2 years). One year actuarial survival for disease groups ranged from 60% for CF to 88% for CHD. Despite these gratifying results, we have identified 6 issues critical to the survival of pediatric lung transplantation, to both the individual patients and to the programs themselves. Our experience and management strategies in these areas are reviewed: CYTOMEGALOV1RUS (CMV): 6/8 (75%) CMV mismatched patients (Donor +/ Recipient -) and 7/32 patients who survived greater than 30 days (23%) developed CMV disease. All but CMV Donor -/ Recipient - patients were treated with ganciclovir for 4 weeks posttransplant. We are presently investigating the use of adjuvant CMV specific immunoglobulin. OBLITERATIVE BRONCHIOLITIS (OB): 5/32 (16%) patients who survived greater than 30 days developed OB. OB was manifest within the first posttransplant year as a rapid decline in small airway function. Aggressive augmentation of immunosuppression has been used with little success. POSTTRANSPLANT LYMPHOPROLIFERATIVE DISEASE (PTLD): 5/32 (16%) patients who survived greater than 30 days developed PTLD. One patient died (17% mortality) despite retransplantation. Four patients resolved their PTLD with reduction in immunosuppression alone, and 1 required the addition of alpha-interferon. We now obtain donor and recipient EBV serology in all lung transplants. CYSTIC FIBROSIS: CF is the leading indication for lung transplantation in children and carries the highest risk, in the form of infection. We have thus changed our management strategies to avoid triple drug immunosuppression, perioperative blood and bronchial cultures, aggressive antimicrobial therapy, and exclusion of patients with panresistant organisms; this has resulted in elimination of infectious mortalities thus far in the pediatric CF group. AIRWAYS: In 21 heart-lung recipients with tracheal anastomoses we have had no airway complications. The double and single lung transplant recipients accounted for 34 bronchial and 1 tracheal anastamoses. There were 3/34 (9%) bronchial stenoses. Two were treated with silastic stents and 1 with balloon dilatation. FINANCES: The average charge for lung transplant evaluation was $18, 000 and for transplantation, $175, 000. The future and success of pediatric lung transplantation will depend upon improved recognition and aggressive prophylaxis and therapy in these 6 critical areas.

*By invitation

29. IMPROVED RESULTS OF LUNG TRANSPLANTATION FOR END STAGE CYSTIC FIBROSIS

Thomas M. Egan, M.D.*, Frank C. Detterbeck, M.D.*, Michael R. Mill, M.D.*, Jeanette T. Thompson, B.S.N.*, Linda J. Paradowski, M.D.* and Benson R. Wilcox, M.D.

Chapel Hill, North Carolina

Although cystic fibrosis (CF) patients (pts) are thought to be at high risk for lung transplantation (LTX) because of the infectious nature of their disease and their nutritional status, our experience would suggest otherwise. Since Oct. 1990, 39 CF pts have undergone double LTX (17 males, 22 females; mean age 23, range 8-45; mean weight 75% expected). Immunosuppression consisted of cyclosporine, azathioprine, antilymphocyte globulin, and prednisone, begun two weeks postop. Three pts required retransplant - 2 for primary graft failure and 1 for obliterative bronchiolitis (OB) at 14 months. All pts were colonized with resistant Pseudomonas (P) aeruginosa, and 6 harbored P cepacia pre-LTX. Six LTX procedures were performed on ventilated pts. Bilateral sequential implantation was used with bronchial omentopexy. Graft ischemic times were 303 ± 9.3 min for the first implanted lung and 456 ± 13.2 min for the second (mean ± SEM). Cardiopulmonary bypass was required for 6 procedures in 5 pts and was associated with increased blood transfusions, compared to LTX performed without it. There have been no operative deaths; three pts died within the first 6 months, two from P cepacia pneumonia and one, unexplained increased intracranial pressure. Of the four pts with partial airway dehiscence, three healed and one was retransplanted. Two pts required temporary dialysis for postoperative renal failure. Hospital stay averaged 30 days (range 14-129 days). One year actuarial survival is 83%. Pre-and postoperative FVC, FEV-1 and actuarial survival of recipients is tabulated (mean ± SEM).

Survivors have returned to an active lifestyle off oxygen. Infection-related morbidity in CF pts is the same as that for pts with other LTX indications. Of 26 pts surviving beyond 6 months, 11 developed OB, which led to late death in 4 pts. Non-compliance contributed to two of these deaths. Other causes of late death include suicide (1), lymphoma (1), and CMV pneumonia after retransplant for OB (1). LTX can be successfully performed in end stage CF pts with acceptable morbidity and mortality.

*By invitation

30. EXPLORATORY ANALYSIS OF TIME-DEPENDENT RISK FOR INFECTION, REJECTION AND DEATH AFTER PULMONARY TRANSPLANTATION

Ko Bando, M.D.*†, Irvin L. Paradis, M.D.*, Robert L. Hardesty, M.D., John M. Armitage, M.D., Kanshi Komatsu, M.D.*, Hiroaki Konishi, M.D.*, Robert J. Keenan, M.D.*, Henry T. Bahnson, M.D. and Bartley P. Griffith, M.D.

Rochester, Minnesota and Pittsburgh, Pennsylvania

Background: Infection and rejection remain the greatest threat to the survival of pulmonary allograft recipients. Furthermore, there is a complex relationship between infection and rejection in these patients since occurrence of one may predispose to the other. Methods and Results: Using a multivariate analysis for repeated events, we analyzed risk factors for treated infection, acute rejection (AR;≥ grade II) and death among 200 recipients who received 210 pulmonary transplants between January 1988 and March 1993. A total of 76 deaths, 432 AR episodes, and 279 distinct infectious episodes occurred during a follow-up of 6 to 69 months. The pattern of AR after transplant was triphasic, characterized by an early period of higher risk (greatest during the first month), a second lower risk period at 2.3 years (midterm), and a low constant risk. The pattern of infection also appeared triphasic, with a delayed, early phase (peak at 3 months post-op), a period of increased risk at 2.5 years post-op, and a late constant phase. By multivariate analysis, risk factors for AR early after transplantation were donor/recipient cytomegalovirus (CMV) mismatch (p=0.0001), longer donor organ ischemia (p=0.001) and older donor age (p-0.01). Risk factors for AR during the midterm and constant hazard phase included longer donor organ ischemic time (p=0.004), symptomatic CMV disease after transplant (p=0.01), greater donor/recipient HLA mismatch (HLA-A, B, or DR)(p<0.01), older donor age (p=0.01), and multiple previous rejection episodes (p=0.03). A risk factor for early infection was an episode of significant AR (≥ Grade II) (p=0.001), while risk factors for late infection included frequent episodes of significant AR (p=0.001), previous symptomatic CMV disease (p=0.001) and frequent episodes of bacterial infection (p=0.01). Multiple regression analysis demonstrated that eventual nonsurviving patients had significantly higher rates of AR and infection during both the early and late phases compared with survivors (p<0.05). The increased rate of AR among nonsurvivors was evident throughout follow-up, although no deaths were attributable directly to AR after the first 9 months. Conclusion: These data suggest that CMV disease, donor ischemic time and donor age affect the incidence of early and late AR. Incidence of infection is increased with repeated episodes of AR. A complex interrelationship between infection and AR determines late survival after pulmonary transplantation.

†1991-1992 Graham Fellow

*By invitation

31. SUCCESSFUL OUTCOME OF LUNG TRANSPLANTATION IS NOT COMPROMISED BY THE USE OF MARGINAL DONOR LUNGS

Sudhir Sundaresan, M.D.*, Janice Semenkovich, M.D.*, Elbert P. Trulock, M.D.*, Laura Ochoa, R.N.*, Greg Richardson, R.N.*, Joel D. Cooper, M.D. and G. Alexander Patterson, M.D.

St. Louis, Missouri

Lung transplantation is limited by a shortage of suitable donors. To address this shortage, we have begun utilizing donor lungs which do not meet our previous rigorous donor criteria. Of 100 consecutive lung transplants performed between June 1991 - August 1993, 68 donor lungs were considered ideal as they satisfied all of the following accepted donor criteria (Group I); Age < 55; Smoking < 20 pack years; PaO₂ > 300 mm Hg (using FIO₂ = 1.0 and positive end-expiratory pressure 5 cm H₂O); and chest radiograph negative for ipsi - or contralateral infiltrate or trauma (contusion or pneumothorax). 32 lungs were considered marginal (Group II) based on criteria shown in the following table:

53 single lung transplants (SLT) were performed (Group I, 37 vs. Group II, 16) compared to 47 bilateral sequential transplants (BSLT; Group I, 31 vs Group II, 16). In 17 cases in Group II, one or both transplanted lungs contained contusion or infiltrate. Recipient outcome, summarized in the following table, did not differ significantly between the groups:

Cardiopulmonary bypass was required to facilitate second graft insertion in 5/31 BSLT in group I (16%) compared to 4/16 BSLT in group II (25%). On the basis of these data, we conclude that successful outcome of lung transplantation can be achieved with the use of marginal donor lungs.

4:45 p.m. EXECUTIVE SESSION (Members Only)

East Ballroom

*By invitation

TUESDAY AFTERNOON, APRIL 26, 1994

1:45 p.m. SIMULTANEOUS SCIENTIFIC SESSION C - CONGENITAL HEART DISEASE

West Ballroom

Moderators: Edward L. Bove, M.D.

William G. Williams, M.D.

32. DOES THE ROSS OPERATION PROVIDE A DEFINITIVE SOLUTION FOR CHILDREN WITH COMPLEX LEFT VENTRICULAR OUTFLOW TRACT OBSTRUCTION?

Jan M. Quaegebeur, M.D., David Solowiejczyk, M.D.*, Daphne Hsu, M.D.*, François Bourlon*, John Hess* and Welton Gersony, M.D.*

New York, New York; Monaco and Rotterdam, The Netherlands

Since 1988, thirty-three children (age 6 weeks - 15 years) underwent aortic root replacement with a pulmonary autograft (Ross operation) and their survival at five years was 100%. The primary lesion was valvar stenosis (12), hypoplastic aortic annulus (7), long segment LVOTO (9) and aortic incompetence (5). The patients with stenotic lesions had undergone previously 10 balloon dilatations and 39 surgical interventions, amongst them 2 Konno operations. In addition to the Ross operation, the LVOT was enlarged with a pericardial patch in 2 patients, and with a septal incision, using the RV muscle of the autograft to augment the septum in 5 patients. Enucleation of fibromuscular obstruction was performed in 4, repair of aortic arch stenosis in 1, and closure of residual VSD in 1 patient. The Ross operation was always done as a root replacement. Aortic and pulmonary allografts were used for RVOT reconstruction.

Most recent post-operative Echo-Doppler studies of the LVOT estimated gradients of 0-10mm Hg in 20 patients, 10-20mm Hg in 2 and 20-30mm Hg in 1 patient. Neo-aortic incompetence was absent or trivial in 30 patients, mild in 2 and moderate in 1. The degree of neo-aortic incompetence in the latter 3 patients was present early postoperatively and has not progressed with time. There were no gradients across the RVOT in 22 patients, 10-20mm Hg in 9, 20-30mm Hg in 1. In 1 patient the gradient was 70mm Hg and he was reoperated for allograft replacement. No other reoperations were required. In 20 patients, serial echo measurements have been obtained. Increase in size of the neo-aortic root was demonstrated in all. In 15 patients this enlargement was appropriate for the patients increase in BSA, supporting growth of the autograft. However, in 5 patients the autograft had become larger than expected for their BSA, suggesting an element of dilatation.

In conclusion, the Ross operation provides excellent relief of complex LVOTO in children of all ages. With appropriate growth of the autograft, in the absence of future structural failure, it appears likely that the Ross procedure is a definitive operation. Autografts which are larger than anticipated will require careful follow-up, and the fate of the allograft in the RVOT remains uncertain.

*By invitation

33. LATE RESULTS OF SYSTEMIC ATRIOVENTRICULAR VALVE REPLACEMENT IN CORRECTED TRANSPOSITION

Jacques A.M. van Son, M.D.*, Gordon K. Danielson, M.D., James C. Huhta, M.D.*, James B. Seward, M.D.*, Hartzell V. Schaff, M.D., Francisco J. Puga, M.D. and Duane M. Ilstrup, M.S.*

Philadelphia, Pennsylvania and Rochester, Minnesota

From 1964 through 1992, 40 patients (aged 5 months to 70 years, median 13.6 years) with corrected transposition of the great arteries and systemic atrioventricular (AV) valve insufficiency underwent replacement (n = 39) or repair (n = 1) of the systemic AV valve. Thirty-nine patients had situs solitus and one had situs inversus. Associated anomalies included Ebstein's malformation of the systemic AV valve (n = 22), ventricular septal defect (n = 10), and pulmonary stenosis (n = 14). Preoperatively, 16 patients (40.0%) had complete heart block and 27 patients (67.5%) were in New York Heart Association (NYHA) functional classes III and IV. The hospital mortality was 10.0% (n = 4), and 8 patients died subsequently. Overall survival was 74.5% at 5 years and 59.3% at 10 years. The principal cause of death was systemic ventricular failure in 12 patients. Survivorship correlated with systemic ventricular ejection fraction of 44% or more (p<0.001) and later interval of operation (9 deaths in 15 patients (60.0%) before 1981 versus 3 deaths in 25 patients (12.0%) subsequently) (p=0.06). There was no surgically-induced complete heart block. Two patients underwent late reoperations related to the systemic AV valve prosthesis. Follow-up extended to 26.0 years (median = 4.3 years). At last follow-up, 18 of the 28 survivors were in NYHA functional class I, 9 were in class II, and 1 was in class III. We conclude that the results of systemic AV valve replacement in corrected transposition have improved significantly during the last decade. In order to preserve systemic ventricular function, operation should be considered at the earliest sign of progressive ventricular dysfunction as assessed by serial clinical evaluation and echocardiography.

*By invitation

34. SURGERY FOR CONGENITALLY MALFORMED MITRAL VALVE IN INFANCY

Miguel Sousa Uva, M.D.*, François Lacour-Gayet, M.D.*, Jaqueline Bruniaux, M.D.*, Alain Serraf, M.D.*†, Jean Paul Binet, M.D. and Claude Planche, M.D.

Paris, France

Congenital mitral valve disease presenting in infancy is rare and surgery is seldom indicated in view of its poor results. However, severe refractory cardiac failure due to mitral incompetence or stenosis, may occasionally require early surgical treatment. The aim of this study was to assess indications and outcome of mitral valve repair in the first year of life.

All patients (pts) less than one year old operated for congenital mitral valve incompetence (MI) (n=10) or congenital mitral stenosis (MS) (n=5) between 1980 and 1993 were retrospectively analysed. Pts with discordant A-V connection, A-V canal defect, cor triatriatum, supravalvular mitral ring or class III/IV hypoplastic left heart syndrome were excluded. Mean age at operation was 6.5 months. Indication was severe heart failure resistant to medical treatment and grade 4/4 MI, or mean pulmonary artery pressure >45 mmHg. Associated DORV with subaortic stenosis (n=2), VSD (n=3), aortic stenosis (n=2) and coarctation (n=2) were previously or concomitantly treated. Valve repair was performed in all 10 MI and 3 MS; valve replacement was requred in two pts with parachute mitral valve. Repair included, among other procedures, Wooler or de Vega type annuloplasty and mobilisation of the subvalvular apparatus.

There were no early deaths. Three MI pts required early reoperation with valve replacement in two and re-repair in one. One patient with MS required valve replacement at 6 months and died. No other late deaths occurred. Follow up has been one month to 10 years (mean=50 months). Eleven pts were in NYHA class I and 3 were in class II. Doppler echocardiography at latest follow up after valvuloplasty was available in 8 pts and showed minimal MI in 3 pts, mild MI in 3 and moderate MI in 2; two pts had a maximum trans mitral gradient of 10 mmHg. These results show that : 1) valve repair for congenital MI in infancy can be performed with satisfactory functional results and low operative risk although some patients will require reoperation. 2) MS remains a surgical challenge, reparative procedures being palliative and valve replacement most often unavoidable.

2:45 p.m. INTERMISSION - VISIT EXHIBITS

†1993-1994 Graham Fellow

*By invitation

3:15 p.m. SIMULTANEOUS SCIENTIFIC SESSION C -CONGENITAL HEART DISEASE

West Ballroom

Moderators: Edward L. Bove, M.D.

William G. Williams, M.D.

35. LONG-TERM FOLLOWUP OF EXTENDED AORTOPLASTY FOR SUPRA VALVULAR AORTIC STENOSIS

Ralph E. Delius, M.D.*, John B. Steinberg, M.D.*, Thomas J. L'Ecuyer, M.D.*, Donald B. Doty, M.D. and Douglas M. Behrendt, M.D.

Iowa City, Iowa and Salt Lake City, Utah

Extended aortoplasty is an operation that was designed to provide a symmetric reconstruction of the aortic root in patients with discrete supravalvular aortic stenosis. The aim of this report is to provide long term followup of the original cohort of patients undergoing this operation. Fifteen patients underwent extended aortoplasty between 1977 and 1983. Followup was obtained in fourteen patients. One patient was lost to followup 3 years after operation; he was included in this report. An echocardiogram, chest radiograph, and electrocardiogram were obtained for each surviving patient. The median length of followup was 141 months (range 36-238). The median preoperative gradient was 90mm Hg (range 55-150). The median immediate postoperative gradient was 20mm Hg (range 0-50, p<0.05 compared to preoperative gradient) and the median long term gradient was 32mm Hg (range 6-96, p<0.05 compared to preoperative gradient, p=NS compared to immediate postoperative gradient). Two patients died during the period of followup, one from left ventricular failure following an aortic valve replacement and one from chronic left ventricular failure. The Kaplan-Meier estimate of survival at 169 months for all patients was 77.4% (70% CL 62-93%). The estimated freedom from reoperation for all patients was 60% at 141 months (70% CL 56-82%). Univariate analysis revealed that the presence of a bicuspid valve is a significant risk factor for reoperation (p=0.038), but not for death (p=0.51). The Kaplan-Meier estimate of freedom from reoperation for patients with a bicuspid aortic valve was 42.9% at 141 months (70% CL 21-65%). Extended aortoplasty provides effective long term relief of the pressure gradient across the supravalvular ridge. However, a significant number of patients require subsequent operations, particularly those with a bicuspid aortic valve.

*By invitation

36. CRITICAL AORTIC STENOSIS: A COMPARISON OF BALLOON VALVULOPLASTY AND TRANSVENTRICULAR DILATATION

Ralph S. Mosca, M.D.*, Mark D. lannettoni, M.D.*, Steven M. Schwartz, M.D.*, Robert H. Beekman, III, M.D.*, Achi Ludomirsky, M.D.* and Edward L. Bove, M.D.

Ann Arbor, Michigan

The optimal treatment of critical aortic stenosis in the neonate and young infant remains controversial. We retrospectively reviewed our experience with transventricular dilatation (TVD) utilizing normothermic cardiopulmonary bypass and balloon valvuloplasty (BV) with respect to early and late survival, relief of aortic stenosis, degree of aortic insufficiency, left ventricular function, and freedom from reintervention. Between July 1987 and July 1993, 24 neonates and infants underwent either TVD (n = 17) or BV (n = 7) for critical aortic stenosis. The patients in the TVD group were older (mean age, 18 days; range, 1-59 days) when compared with the BV group (mean age, 10 days; range, 1-31 days), p=.05. There were no significant differences in left ventricular function, degree of left ventricular outflow tract obstruction, or aortic annulus size between the groups. Associated anomalies were more common in the TVD group (53%) compared to the BV group (13%), p=.05. Following treatment, the mean reduction in aortic gradient and the degree of aortic insufficiency were equivalent in both groups as assessed by Doppler/echocardiography. Ejection fraction improved significantly in both groups (TVD, 39 ± 5.0% vs 46 ± 5%; BV, 51 ± 2% vs 62 ± 3%) but there was no statistical difference between groups. The LV mass to volume ration (gm/cc³) also increased significantly in both groups but with no significant difference between groups (TVD, 1.2 ± 0.5 vs 1.6 ± 0.6; BV, 1.1 ± 0.6 vs 1.5 ± 0.4). Early mortality in the TVD group was 9.5% and in the BV group 14% (p=ns). There were no late deaths in either group. Three patients from the TVD group and 2 patients from the BV group required reintervention for further relief of aortic stenosis. We conclude that both TVD and BV provide adequate and equivalent relief of critical aortic stenosis. Treatment strategies should therefore depend upon other factors including the presence of associated cardiovascular anomalies, vascular access, preoperative condition and the local expertise of the institution.

*By invitation

37. COARCTATION OF THE AORTA PRESENTING FOR TREATMENT IN NEONATES

John W. Kirklin, M.D., Jan M. Quaegebeur, M.D., Richard A. Jonas, M.D., Alan D. Weinberg, M.S.* and Eugene H. Blackstone, M.D.

Birmingham, Alabama; New York, New York and Boston, Massachusetts

Among the 322 neonates with coarctation, with or without a ventricular septal defect (VSD) and in a multi-institutional study, survival for at least 24 months after an initial procedure was 84%. Other coexisting obstructive lesions in the left heart-aorta (LHA) complex decreased survival; these included mitral valve anomalies in 5% of patients, left ventricular hypoplasia in 5%, narrowing of the left ventricular outflow tract in 9%, and narrowing of the proximal arch in 1%. The most commonly used technique of repair of the coarctation was resection and end-to-end anastomosis, but no technique was a risk factor for death by multivariable analysis. Extension of the area of resection so that the end-to-end anastomosis was proximal to the left subclavian artery but distal to the left common carotid artery, did not increase risk. Neonates without other coexisting obstructive lesions in the LHA complex, and without or with small VSD, have a 97% survival for at least 24 months after end-to-end anastomosis, or subclavian flap repair, or patch graft repair; among those with coexisting moderate-sized or large VSDs, repair of the VSD and pulmonary trunk (PT) banding was associated with the highest 2-year survival, 97% in those with single VSD (Figure).

The risk-adjusted outcomes in two institutions were believably less; good than in all others. Therapeutic inferences are that critically ill neonates with coarctation without or with small VSD do well with surgical repair of the coarctation as initial treatment. Those with moderate-sized or large VSD do best with initial repair of the coarctation and PT banding (often with subsequent repair of the VSD at the same hospitalization). Concomitant procedures against a coexisting obstructive lesion in the LHA complex are rarely indicated.

*By invitation

38. VIDEO-ASSISTED THORACOSCOPIC SURGERY FOR CONGENITAL HEART DISEASE

Redmond P. Burke, M.D.*, Mary VanderVelde, M.D.*, Dolly Hansen, M.D.* and Gil Wernovsky, M.D.*

Boston, Massachusetts

Video-assisted thoracoscopic surgery (VATS) provides excellent visualization of anatomic structures, causes minimal surgical trauma, and has significantly expanded surgical options in adult thoracic surgery. VATS applications in children with congenital heart disease have been limited to patent ductus arteriosus interruption. After designing endoscopic instruments for pediatric thoracoscopic cardiovascular use, and extensive animal experimentation, video-assisted techniques were developed for 7 different surgical procedures in children, including: Interruption of patent ductus arteriosus (N=14), division of vascular ring (N=4), drainage of posterior pericardial effusion (N=3), interruption of arterial and venous collaterals (N=2), thoracic duct ligation (N=1), epicardial pacemaker lead insertion (N=1), and diagnostic thoracoscopy (N=1). Ages of these 26 patients ranged from 2 hours to 12 years, and weights from 575 grams to 46 kilograms. There was no operative mortality. Five patients (19%) required conversion to thoracotomy to complete procedures including: vascular ring division (N=2), pericardiectomy (N=1, PDA interruption (N=1) and epicardial pacemaker lead placement (N=1). One recurrent laryngeal nerve injury (4%) occurred during PDA interruption. Patients undergoing elective VATS for vascular ring division and PDA ligation were extubated in the operating room and discharged from the hospital within 48 hours. With a comprehensive program, VATS can be successfully applied to a widening range of cardiovascular problems afflicting neonates and infants.

4:45 p.m. EXECUTIVE SESSION (Members Only)

East Ballroom

*By invitation