WEDNESDAY MORNING, May 10, 1989

7:30 a.m. SYMPOSIUM I - AORTIC DISSECTION

MODERATOR: Randall B. Griepp, M.D., New York, NY

PANELISTS: E. Stanley Crawford, M.D., Houston, TX

M. Arisan Ergin, M.D., New York, NY

Nicholas Kouchoukos, M.D., St. Louis, MO

D. Craig Miller, M.D., Stanford, CA

7:30 a.m. SYMPOSIUM II - SURGERY OF EMPHYSEMA

MODERATOR: Jean DesLauriers, M.D., Sainte-Foy, Quebec, Canada

SURGICAL APPROACH TO BULLOUS EMPHYSEMA

Edward A. Gaensler, M.D., Boston, MA

OPERATIVE STRATEGIES AND RESULTS OF SURGERY

Jean-Paul Whitz, M.D., Strasbourg, France

ROLE OF SURGERY IN DYSPLASTIC TRACHEA

Mr. H. Hsrzog, Basel, Switzerland

NEW PERSPECTIVES IN SURGERY FOR DIFFUSE NON BULLOUS EMPHYSEMA

Marcel Dahan, M.D., Toulouse, France

9:00 a.m. SCIENTIFIC SESSION - HYNES BALLROOM

26. 30 Year Follow-Up of Superior Vena Cava-Pulmonary Artery Shunts

GARY S. KOPF*. HILLEL LAKS,

HORACE C. STANSEL, WILLIAM E. HELLENBRAND*,

CHARLES S. KLEINMANN* and NORMAN S. TALNER*

New Haven, Connecticut

Superior vena cava-pulmonary artery shunt (SVC-PAS), (Glenn shunt) as first performed clinically at our institution in 1958. From then through September 1988, 91 patients have undergone SVC-PAS. We here report the follow up data available on all patients.

Patients' age ranged from 1 month to 46 years (mean 6.8 yrs). Diagnoses were: tricuspid atresia 27, single ventricle 22, tetralogy of Fallot 14, D-TGA +VSD + PS 9, TGA 5, Ebstein's anomaly 4, PA + intact septum 4, others 6. Hospital mortality was 7.7% (1 death in the last 53 patients 1.9%); 5 of the patients who died were 4 months of age or younger.

Arterio-venous fistula formation (AVF) occurred in 18 patients (19.7%), 6 of whom have undergone therapeutic embolization with improvement in saturation. The incidence of AVF increased with time post shunt. Desatura-tion due to venous collaterals occurred in 2 patients, and a patent SVC-RA connection in 1. No shunt thrombosis or obstruction due to stricture formation occurred following discharge from the hospital. Improvement in saturation was obtained following SVC-PAS in 8 inoperable patients by creation of a right axillary arterio-venous fistula up to 12 years following shunt placement. Three patients had conversion of a right Blalock-Taussig shunt to a SVC-PAS to decrease the volume load on a failing ventricle and to improve saturation. Thirty patients underwent physiological repair (Fontan procedure or modification in 26 and biventricular repair in 4, with 4 deaths (13.3%). In 3 instances, the SVC-PAS was taken down to provide continuity between the left and right pulmonary arteries; otherwise the shunt was left intact. Compared to patients without prior SVC-PAS, hospital mortality was not significantly different. Thirty years following SVC-PAS, the first patient in this series is married and working full time, having undergone a modified Fontan procedure (RA-LPA connection) in 1981.

We conclude that SVC-PAS, usually with supplemental procedures to enhance oxygenation, has provided excellent physiologic palliation with low mortality up to thirty years with no late thrombosis or stricture formation. Pulmonary A-V fistula formation increases with embolization. Physiological repair following SVC-PAS carries a low mortality with good long term survival. Although currently used infrequently, SVC-PAS, particularly with end-to-side anastomosis, remains a useful method of palliation in selected patients.

*By Invitation

27. The Bi-Directional Cavopulmonary Shunts

JOHN J. LAMBERTI, ROBERT L. SPICER*,

TODD M. GREHL* and J. DEANE WALDMAN*

San Diego, California

The bi-directional cavopulmonary shunt (CPS) improves systemic arterial oxygen saturation without increasing ventricular work or pulmonary vascular resistance. A CPS is ideal palliation for patients destined for a right atrium to pulmonary artery connection (RH-PAC). Since 1983, 15 patients have undergone CPS (5 primary operations, 10 secondary operations). Diagnoses were: asplenia; 2, single ventricle; 2, pulmonary atresia and intact ventricular septum; 6, tricuspid atresia; 3, hypoplastic left heart; 1, TGA with hypoplastic LV; 1. Age at primary operation ranged from 3.5 to 30 months (median = 6 mos.). Weight ranged from 3.5 kg. to 9.7 kg. Age at secondary operation ranged from 10 mos. to 14 yrs. (median = 15 mos.). Six CPS were performed through a right thoracotomy utilizing a temporary shunt. Nine CPS were performed on cardiopulmonary bypass (CPB). All CPB patients had additional procedures: Takedown modified Blalock-Taussig shunt, 7; revise right ventricular outflow tract, 4; reconstruct pulmonary arteries, 4; tricuspid valvuloplasty, 1. There was no operative mortality. One patient required early revision. Follow-up ranges from 1 to 53 months. Twelve of 15 had a good to excellent late result. Two patients died late (pulmonary vascular disease, 1 pt., pulmonary A-V malformations, 1 pt.). There was one late failure (converted to Glenn). The CPS is an excellent palliative procedure when RA-PAC must be deferred because of age, weight or anatomic considerations. In addition, at the time of RA-PAC (modified Fontan) the CPS approach may optimize the anatomic connection (8 additional patients).

*By Invitation

28. Unifocalization and Complete Repair of Patients with Pulmonary Atresia, Ventricular Septal Defect and Systemic Collaterals

FRANCISCO J. PUGA, FRANCO E. LEONI*,

PAUL R. JULSRUD* and DOUGLAS D. MAIR*

Rochester, Minnesota and Edmonton, Alberta, Canada

Arborization abnormalities of the peripheral pulmonary arterial tree in patients with pulmonary atresia and ventricular septal defect have limited the success of attempts at complete surgical repair. Techniques at surgical Unifocalization have allowed disconnection of pulmonary arterial branches from systemic collateral arteries with concomitant anastomosis to other branches and creation of centrally accessible sources of pulmonary arterial blood flow. Thus, an anatomic situation is created which permits complete correction of the anomaly. From 1982-1988 we have operated 115 patients in whom 60 unifocalizations were performed. All patients had pulmonary atresia, ventricular septal defects, and systemic collateral arteries supplying the majority of pulmonary arterial segments. A total of 93 extrapericardial anastomoses were performed and 80 systemic collaterals interrupted. Procedures included: direct pulmonary arterial branch anastomosis (Type I), in-terpulmonary conduit interposition (Type II), patch angioplasty of communicating pulmonary arterial branches (Type III), and establishment of in-trapericardial pulmonary confluence (Type IV). There were 3 deaths resulting from the Unifocalization procedure (5%). Twenty-five patients have completed bilateral Unifocalization procedures. Two of these have been rejected for complete repair due to persistent restriction in pulmonary arterial runoff. One patient has been accepted for complete repair and is waiting for surgical correction. Twenty-two patients have undergone complete repair (elimination of extracardiac sources of pulmonary arterial blood flow, closure of septal defects, and insertion of valved conduit between the right ventricle and the pulmonary arteries). There were 21 survivors. Intraoperative, post-repair peak systolic pressure ratio between the right and left ventricles ranged from 0.4 to 0.85 (mean = 0.60, SD 0.25). We conclude that, in selected patients, staged reconstruction of peripheral arterial confluence may correct arborization problems sufficiently to allow complete surgical repair of patients formerly considered unrepairable.

10:00 a.m. Intermission - Visit Exhibits

*By Invitation

10:45 a.m. SCIENTIFIC SESSION - HYNES BALLROOM

29. The Size of the Pulmonary Arteries and the Results of the Fontan Type Repair

FRANCIS FONTAN, GUY FERNANDEZ*,

FRANCISCO COSTA*, DAVID NAFTEL*,

FRANCESCO TRITTO*, EUGENE BLACKSTONE

and JOHN W. KIRKLIN

Pessac-Bordeaux, France, Curitiba, Brazil

and Birmingham, Alabama

Among the 334 patients undergoing a Fontan type repair for congenital heart disease (between 1968 and August 1, 1988) in two widely separated institutions, cineangiograms were available in 234 patients (median age 7.0 years, range 8/12 to 38 years) for measurements of the McGoon ratio:

diameter of right pulmonary artery + diameter of left pulmonary artery

diameter of descending aorta at diaphragm

Fifty-five patients (24%) died or had a takedown of the repair within two months of the operation (the event). 124 patients had tricuspid atresia (20 events) and 110 had other cardiac malformations, most commonly double inlet left ventricle (35 events) (P = .005). The new atrial connection was to the right ventricle in 73 patients (8 events) and to the pulmonary arteries in 161 (47 events) (P = .002). The median value of the McGoon ratio was 2.3. A smaller McGoon ratio was a risk factor (logistic multivariate analysis) for death or takedown within two months of the Fontan-type repair (P<.001) with increments being added by young age (P<.001), termination of the new right atrial connection on the pulmonary artery (rather than right ventricle) (P<.007) and the presence of mitral atresia (P = .03). The risk factors were identical when the event was analyzed in a time-related manner (hazard function). Solution of the multivariate equation, in a patient aged 5 years, without mitral atresia, and with a right atrial to right ventricular connection, predicted a 14%, 22%, and 41% prevalence of the event when the McGoon ratio was 2.0, 1.6 and 1.2 respectively; and when the connection was directly to the pulmonary arteries, the prevalence was 36%, 49% and 71% respectively.

To enhance usefulness in considering early takedown of the Fontan repair, an additional multivariate analysis was made incorporating the early postoperative right atrial pressure (P_RA)- This considerably refined the risk factors, such that in the patient described, with a connection to the right ventricle and with a McGoon ratio of 2.0, the prevalence of the event varied from 5% (P_RA 13 mmHg) to 15% (P_RA17 mmHg) to 38% (P_RA 21mm Hg); with a McGoon ratio of 1.6 the prevalence of the event varied from 10% (P_RA 13 mmHg) to 27% (P_RA 17 mmHg) to 56% (P_RA21 mmHg); and with a McGoon ratio of 1.2 the prevalence of the event varied from 27% (P_RA 13 mmHg) to 55% (P_RA 17 mmHg) to 81% (P_RA21 mmHg). (These combinations were all represented in the experience.)

Therefore, the size of the right and left pulmonary arteries, as determined by preoperative cineangiography, is an important risk factor for surgical failure, but its effect can be modified by other risk factors.

*By Invitation

30. Modified Fontan Procedure Reconstructive Surgery for Single or Dominant Right Ventricle

JOHN D. PIGOTT*, ALVIN J. CHIN*,

JOHN D. MURPHY* and WILLIAM I. NORWOOD

Philadelphia, Pennsylvania

Since January, 1984, 105 patients with a single or dominant right ventricle have undergone application of Fontan's procedure for definitive palliation. Diagnosis included hypoplastic left heart syndrome (60), heterotaxy syndrome (23), solitary right ventricle (4), other (18). Age at Fontan's procedure ranged from 4.5 months to 23 years (median 20 months). Surgery included baffling pulmonary venous return in 15 patients, closure of an ASD and atriopulmonary anastomosis in 6 patients, and baffling systemic venous return to the pulmonary arteries in 84 patients. Extensive pulmonary arterial augmentation using pulmonary artery homograft was performed in the last 85 patients. There were 26 early deaths. Suspected causes of early mortality include ventricular diastolic dysfunction (10), residual pulmonary arterial stenosis (4), pulmonary venous hypertension (3), elevated pulmonary vascular resistance (2), hemorrhage and infection (3 each) and hepatic dysfunction (1). Thirty-five patients had 37 nonfatal complications, including prolonged pleural and pericardial effusions (30), complete heart block (3), prolonged ventilation and atrial arrhythmias (2 each). Early reoperation was undertaken in 13 patients: pulmonary artery angioplasty (4), change pulmonary venous to systemic venous baffle (3), pacemaker (2), others (6). Five of 26 early deaths occurred in patients undergoing early reoperation. There were 8 late deaths. Suspected causes of late mortality include persistent pleural and pericardial effusions (4), ventricular diastolic dysfunction (2), irreversible cerebral injury (1), elevated pulmonary vascular resistance (1). Patients with single or dominant right ventricles have structural issues not present in patients with single left ventricles. The thrust of this work has been to develop a systematic means to apply Fontan's principle to all patients with single ventricle, including those with unusual pulmonary and systemic venous return and right atrioventricular valve dysfunction. Important elements for improved survival in this series include: wide augmentation of branch pulmonary arteries and baffling of systemic venous rather than pulmonary venous return at initial operation as well as scrupulous avoidance of elevations in ventricular end diastolic pressure in the early postoperative period.

*By Invitation

31. Primary Repair of Tetralogy of Fallot in Infancy

GILLES D. TOUATI*, PASCAL R. VOUHE*,

PHILIPPE POUARD*, FRANCINE LECA*,

ANTONIO AMODEO* and JEAN-YVES NEVEUX*

Paris, France

Sponsored by: Aldo R. Castaneda, Boston, Massachusetts

From June 1983 to April 1988, 100 consecutive symptomatic infants with Tetralogy of Fallot (without pulmonary atresia) were operated upon. Age ranged from 0.5 to 12 months (mean = 7.3 ± 3.7). Twenty patients were 0.5 to 3 months, 21 were 3 to 6 months and 59 were 6 to 12 months of age. Mean weight was 6.5 Kg ± 1.7.

Seventy patients received a transannular patch. Only 16 patients had hypothermic circulatory arrest, all others had conventional cardiopulmonary bypass.

Hospital mortality was 3%; there were no late deaths; cummulative follow-up was 186 patients/years. Causes of deaths include: hypoplastic pulmonary arteries (4 and 5 months old) and right ventricular failure (4 months old).

The last 48 patients were operated on without mortality; during this period, operative management differed in: 1.) blood cardioplegia repeated every 20 minutes and 2.) ultrafiltration was added to bypass.

The predicted 30-day survivorship after repair was 96-98% (CL 70%) and was 90-99% (CL 95%). No ventricular arrythmias have so far been detected after repair (mean of follow-up = 21.6 months) and echocardiographic ventricular diameter ratios were: 0.60 ± 0.10.

These early results encourage continued primary repair of symptomatic infants with Tetralogy of Fallot thanks to improved surgical and anesthetic management.

*By Invitation

32. Neonatal Aortic Stenosis

KEVIN TURLEY, EDWARD L. BOVE,

JOSEPH J. AMATO, MARK IANNETTONI*

and JOHN YEH*

San Francisco, California, Ann Arbor, Michigan

and Newark, New Jersey

Aortic stenosis in the neonate has in the past been associated with a high surgical mortality. As a result, in the era of percutaneous balloon valvuloplasty, the optimal mode of therapy remains controversial. An approach of metabolic stabilization using cardiopulmonary bypass (CPB) followed by relief of left ventricular outflow tract obstruction (LVOTO) was employed by three institutions, and the results are presented.

During the period 1983-1988 (the valvuloplasty era), 33 neonates with isolated aortic stenosis, patent ductus arteriosus, and/or coarctation of the aortic underwent operative repair. Ages ranged from 1 to 30 days, median 11 days including 16 in the first week of life. There were 25 males and 8 females; and weights ranged from 2.5 to 5.5 kg., mean 3.7 kg. Preoperative conditions included: congestive heart failure in 32, mitral regurgitation in 14, and left ventricular/aortic gradients ranged from 15 to 130 mmHg. Operative therapy included the use of CPB to provide metabolic stabilization in all thirty-three prior to relief of LVOTO. There were 25 open valvulotomies and eight transventricular dilatations. Hospital survival was 85% (28/33) with n.s. difference between the methods (7/8, 21/25). There have been four reoperations with one late death (MVR-Apical-Aortic conduit) and one late sudden death.

The use of CPB for resuscitation of neonates with critical aortic stenosis combined with relief of LVOTO can result in a high operative and late survival. These results support the concept that this technique provides a milieu in which the nenonatal myocardium can optimally respond to relief of obstruction. The results of this technique are a standard against which closed methods such as percutaneous valvuloplasty should be compared.

12:00 p.m. Adjourn for Lunch

*By Invitation

1:30 p.m. SCIENTIFIC SESSION - HYNES BALLROOM

33. Aortic Valve Replacement with Stentless Porcine Aortic Bioprosthesis

TIRONE E. DAVID, CHARLES POLLICK*

and JOANNE BOS*

Toronto, Ontario, Canada

Fatigue tests indicate that the best stent for the aortic valve leaflets is the aortic root. That is probably why hand-sewn aortic valve homografts are more durable than stent-mounted homografts used for aortic valve replacement. Artificial stents not only shorten the durability of aortic valve bioprostheses but also impair their hemodynamic performance.

We have initiated a clinical trial on aortic valve replacement using a Stentless porcine aortic valve which is processed in the same manner as the Hancock II bioprosthesis (fixed with glutaraldehyde under very low pressure and treated with sodium dodecyl sulphate to retard calcification). This bioprosthesis has been implanted in 20 patients. To evaluate its hemodynamic performance, these 20 patients were matched for age, valve lesion, body surface area and bioprosthetic size with 20 cohorts who underwent aortic valve replacement with Hancock II bioprosthesis. The hemodynamic performance of the bioprostheses was assessed by Doppler echocardiography. The results were the following:

This Stentless porcine aortic valve has lower peak systolic gradient and greater effective orifice than its stented version in every bioprosthetic size. The steneless porcine aortic valves, like the hand-sewn aortic valve homografts, are expected to be more durable than the stended porcine aortic bioprostheses.

*By Invitation

34. Prosthetic Ring Mitral Valve Repair: The Second Decade

ALAIN DeLOCHE*, VICTOR A. JEBARA*,

PATRICK M. PERIER*, JOHN Y. M. R ELL AND*,

GILLES D. DREYFUS* and ALAIN F. CARPENTIER

Paris, France and Beirut, Lebanon

The 195 consecutive patients (pts) having undergone prosthetic ring mitral valve repair (PRMVR) between 1972 and 1979 in our institution were reviewed in order to assess long term function of this method of repair.

Patients age ranged from 18 years to 79 years (mean age 48.7 years). There were 107 (55%) males and 88 (45%) females. Mitral valve incompetence was secondary to degenerative valvular diseases in 113 cases (58%), rheumatic disease in 74 cases (38%), ischemia and various other causes in 8 cases (4%). 188 patients (97%) were in NYHA class III and class IV preoperatively and 94 (48%) had atrial fibrillation. The patients were divided into 3 functional groups:

The techniques used include prosthetic ring annuloplasty (187), partial leaflet resection (158), chordae shortening (89), leaflet mobilization (10) and papillary muscle reimplantation {1). Long term follow-up was available in 189 patients (96.8%) for a rate of 2316 patients/year. The actuarial survival at 15 years was 72.2 ± 4.4%. The occurrence of complications were analyzed in linear and actuarial terms:

Half of the reoperations were necessary in the first postoperative year and were proved to be due to technical errors. Among the 157 survivors, 117 (74%) were in NYHA class I and class II and 105 (66%) were in sinus rhythm. Risk factor analysis did not show any statistically significant difference in the long term results with regards to age, etiologies and functional groups and lesions.

We conclude that PRMVR is associated with a low risk of complications and mortality when compared to other alternatives. The technique is reliable and provides stable long term results.

*By Invitation

35. Comparative Clinical Assessment of Mitral Valve Replacement With and Without Chordal Preservation

HANI A. HENNEIN*, JULLE A. SWAIN*,

CHARLES L. McINTOSH*, CHRISTOPHER D. STONE*,

ROBERT O. BONOW* and RICHARD E. CLARK

Bethesda, Maryland

Left ventricular (LV) function often deteriorates both acutely and chronically after mitral valve replacement (MVR) for pure mitral regurgitation (MR). Disruption of the mitral apparatus at operation has been proposed as a major cause of postoperative dysfunction. The hypothesis tested in this clinical study was that MVR with chordal preservation results in more favorable postoperative LV function. Forty-eight consecutive patients with pure MR were studied before and six months after operation by treadmill exercise testing, catheterizations, echocardiography, and radionuclide angiography. Thirteen patients underwent MVR with St. Jude and Biopros-thesis with preservation of chordae tendineae, and the remaining 35 had MVR without chordal preservation. Preoperatively, there were no differences in age, gender, New York Heart Association functional class, exercise capacity, cardiac index, rest EF, exercise EF, fractional shortening, mean arterial pressures, or cardiac index the two groups. There were three operative and six late deaths among the 35 patients with chordal resection, but no early or late deaths in patients who had MVR with chordal preservation (P = 0.05). In patients whose chordae were excised, exercise capacity was unchanged after MVR and left ventricular function deteriorated with a reduction in resting EF (49 ± 11% to 34 ± 13%, PC 0.001), exercise EF (51 ± 14% to 38 ± 16%, PC 0.01), and fractional shortening (34 ± 10% to 23 ± 18%, p = 0.005). In contrast, exercise capacity increased after MVR in patients whose chordae were preserved (by 5.9 ± 2.2 minutes, PC 0.005) and function was maintained, with no charges in resting EF (45 ± 13% to 47 ± 11%,P = NS), exercise EF (47 ± 12% to 51 ± 14%, P = NS), and a decrease in LV end systolic dimension (from 40 ± 8% to 35 ± 6%, P < 0.05). Both group of patients demonstrated improvement in pulmonary artery pressure, mean left atrial pressure, LV end diastolic pressure, cardiac index, with no statistical differences between groups. These data demonstrate that patients who undergo MVR with chordal preservation have improved postoperative survival, exercise capacity, and LV function compared to patients in whom the chordae were excised.

*By Invitation

36. Mitral Valve Repair: Results and Decision Making Process of the Reconstruction

ARRIGO LESSANA*, CARMINE CARBONE*.

MAURO ROMANO*, EVELYNE PALSKY*,

YU HONG QUAN* and GENEVIEVE LUTFALLA*

Aubervilliers and Paris, France and Beijin, China

From 1975 to 1988, 266 patients (pts) underwent mitral valve repair (MVR) procedures for pure or predominant mitral regurgitation (MR). The cause of MR was rheumatic in 176 pts (mean age: 28.4 ± 1.2, x ± SEM) and degenerative in 78 pts (mean age: 54.2± 1.6). Fifty-five percent of the pts were in NYHA class III and IV prior to surgery. Intraoperative assessment of the MV led us to identify 3 major mechanisms of MR: 1) Restriction of leaflet motion by fibrosis (56 pts, Group I); 2) Enhancement of leaflet motion by leaflet and chordae extension and prolapse (137 pts, Group II) and 3) Combination of both (65 pts, Group III). Only 8 pts had isolated dilatation of the annulus. One hundred and forty-nine pts had isolated MR and 117 had associated aortic and/or tricuspid valve or coronary disease. To illustrate more comprehensively how surgical procedures were selected according to the type of valve disease, a 7-min motion pictures has been prepared and will illustrate the different steps of the decision-making process and the subsequent surgical maneuvers. Hospital mortality was 3.8% ± 2.2. Follow-up was 95.9% complete and totaled 1,182.86 patient-years. The postop assessment included clinical examination, and echodoppler study. At 13-year follow-up, survival was 91.9% in Group I, 89.2% in Group II and 96.5% in Group III. The percent freedom from reoperation was 69.8% in Group I, 81.6% in Group II and 64.2% in Group III; the percent freedom from emboli was 97.2%, 97.6%, and 92.5%, respectively. In the group of pts with isolated MR, the results were as follows:

These results suggest that conservative surgery should be used with caution in Group I and III pts. In contrast MVR should be extended in Group II pts. This finding is clinically relevant, since, in Western countries, valve prolapse tends to be a major cause of MR.

*By Invitation

37. Five Hundred and Thirty Patients Undergoing Tricuspid Valve Surgery: 25 Year Assessment Early and Late Phase Hazards and Events

LYNN B. McGRATH*, LORENZO GONZALEZ-LAVIN,

BRIDGETM. BAILEY*, GARY P. GRUNKEMEIR*,

JAVIER FERNANDEZ, FRANCIS P. SUTTER*

and GLENN W. LAUB*

Browns Mills, New Jersey

From January 1, 1961 through December 31, 1987, 530 patients underwent an intracardiac repair which included tricuspid valve surgery. Three hundred fifty-one patients had tricuspid valve repair (66.2%) and 179 had tricuspid valve replacement (33.8%). Mean age at repair was 56.9 years, range 10 years to 79 years. Two hundred and fifty-six patients (48.3%) had undergone previous cardiac surgery. Preoperative variables predicting the requirement for tricuspid valve replacement included: ascites (p = 0.02), hepatomegaly (p = 0.002), pulsatile liver (p = 0.01), peripheral edema (p = 0.01), tricuspid stenosis (p = 0.001), previous cardiac operation (p = 0.001), increasing right atrial pressure (p = 0.0001), increasing preoperative angiographic severity of tricuspid valve incompetence (p = 0.003), concomitant mitral valve replacement (p = 0.002), and higher functional class (p = 0.07). There were 76 hospital deaths (14.3%). Incremental risk factors for the event hospital death included male gender (p = 0.003), higher functional class (p = 0.07), hepatomegaly (p = 0.02), ascites (p = 0.04), previous operation (p = 0.009), total cardiopulmonary bypass time (p = 0.00001), total aortic occlusion time (p = 0.03), and the use of an annuloplasty ring in the tricuspid valve repair group (p = 0.05). Ninety-seven percent of the patients were followed, at a mean of 50.6 months, range 0 to 315 months. There were 250 late deaths (55.1%). Actuarial survival was 33.5% at 120 months and 13.1% at 180 months. Seventy-eight patients (17.2%) had at least one reoperation. Actuarial freedom from reoperation was 83.1 % at 60 months and 29.3% at 180 months. There was no difference in reoperation rates (p = 0.10), nor actuarial survival (p = 0.40) according to the type of tricuspid valve procedure performed. Hazard function analysis for reoperation revealed early, intermediate, and late phase events.

We conclude that preoperative variables may predict the requirement for tricuspid valve replacement. As there is no difference in hospital nor late survival in patients undergoing tricuspid valve repair (versus replacement), and as there is an increased incidence of valve related events in the replacement group, we recommend tricuspid valve repair, without annuloplasty ring, whenever possible.

*By Invitation

38. The Operative Treatment of Patients with Hypertrophic Cardiomyopathy and Pulmonary Hypertension

CHRISTOPHER D. STONE*, HANI A. HENNEIN*,

CHARLES L. McINTOSH*, ARSHED A. QUYYUMI*

and RICHARD E. CLARK Bethesda, Maryland

The clinical course and hemodynamic results in patients undergoing operation for obstructive hypertrophic cardiomyopathy (HCM) with preoperative pulmonary artery hypertension (PHT) was unknown. The hypothesis tested in this retrospective study was that operative relief of left ventricular outflow tract obstruction results in a substantial reduction in pulmonary artery pressures and mitral regurgitation (MR) without necessitating mitral valve replacement (MVR). Patients were included if their preoperative pulmonary systolic pressure was > 35 mmHg, and were, with the exception of MR, without concomitant cardiac pathology. Since 1962, 49 pts. who fit our criteria underwent left ventricular myotomy and myectomy (LVM&M) with 98% follow up. Mean follow up was 7.9 ± 4.5 (mean ± 1 S.D.) years with a range of .8 to 18.4 years. Early hospital mortality was 12% (N ± 6), 2 (4%) from low C.O. and 4 (8%) from arrhythmia. There were 43 (84%) hospital survivors and 19 late deaths; 4 (9%) from arrhythmia or sudden death, 5 (12%) from CHF, 3 (7%) died at repeat cardiac procedure, and 4 (9%) from other causes. Actuarial survival was 74 ± 6.5% (N = 31) at five years and 46 ± 8.4% (N = 8) at ten years. Three patients had late MVR. Of the 43 survivors 39 (91%) returned 7.7 ± 4.5 months later for follow up evaluation including cardiac catheterization. The majority (82%) experienced New York Heart Association functional class one or two status postoperatively. Cardiac catheterization indicated a fall in PA systolic pressure from 62 ± 17.4 (range = 36 to 105) to 38.5 ± 11.7 (range = 17 to 68) mmHg (p = .0001) with no difference in right atrial pressure or cardiac output. Pulmonary capillary wedge mean pressure decreased from 23.6 ±6.4 to 16.1 ± 5.3 mmHg (p = .0002) and, preoperative MR improved or was abolished in 85% of patients studied (N = 13). Rest and maximum provocable left ventricular outflow tract gradients decreased from 80.5 ± 45.4 and 103.4 ± 26.2 to 13.8 ± 19.2 and 46 ± 43 respectively (p = .0001). Analysis of the above pts. operated upon since 1982 and comparison to an age and functionally matched group with PHT and HCM treated with MVR showed no statistical difference in mortality, morbidity, hemodynamic or functional outcome with two years of mean follow up. We conclude that a consistent, significant reduction (mean =42% ± 18%) in preoperative pulmonary hypertension, clinical symptoms and mitral regurgitation occurs with relief of outflow tract obstruction by LVM&M and that MR and PHT are not indications for MVR in these patients.

*By Invitation

SCIENTIFIC SESSION - ALTERNATE

Hypoplastic Left Heart Syndrome: Palliation Without Cardiopulmonary Bypass

WILLIAM Y. TUCKER*, ROBERT C. McKONE*,

KENNETH M. WEESNER* and NEAL D. KON*

Winston-Salem, North Carolina

Sponsored by: A. Robert Cordell, Winston-Salem,

North Carolina

Hypoplastic left heart syndrome (HLH) continues to be a surgical challenge. Mortality with a first stage palliative repair utilizing car-diopulmonary bypass (CPB) is high. Cardiac transplantation in neonates with HLH has severe limits of time, donor avilability and patient management.

Our 100% mortality with first stage HLH repair using CPB led us to a procedure not requiring CPB. In 9 consecutive patients, after medical stabilization, a left thoracotomy was performed and an 8 mm woven Dacron graft placed from the main pulmonary artery to the descending thoracic aorta. The patent ductus arteriosus was ligated and the main pulmonary artery banded distal to the graft and proximal to the bifurcation.

Age at surgery ranged from 2-6 days. Weights were 2.3-3.7 kg (mean 3.2 kg). Pre-op diagnosis was made by ECHO in 4 patients and by cath in 5. Caths were accompanied by balloon atrial septostomy. Two patients diagnosed by ECHO underwent post-op cath and balloon septostomy prior to discharge. Five patients were extubated within 4 days (range 1-4). Only low dose Dopamine was needed for inotropic support. There were no bleeding problems. Five patients were discharged 11-80 days postop (mean 38 days). There were 4 hospital deaths; 1 from Candida sepsis at 81 days, 1 from low cardiac output at 2 days, and 2 from a restrictive ASD at 3 and 5 days. Both of the latter had pre-op ECHO with cath and balloon septostomy planned prior to discharge. Patients followed up to 3 years have had normal growth and development.

Palliation of HLH without CPB has allowed hospital discharge in 5 of 9 patients and has decreased post-op hemodynamic, respiratory and bleeding problems. We now do pre-op balloon septostomies on all patients and feel that this will improve survival rate. This simpler initial approach may allow for a staged Fontan or cardiac transplant at a later date for more definitive treatment.

3:30 p.m. ADJOURN

*By Invitation