37. Critical Aortic Stenosis in the Neonate: a Multi-Institutional Study of Management, Outcomes and Risk Factors

Gary K. Lofland*, Brian W. McCrindle*, William G. Williams, Eugene H. Blackstone, Christo Tchervenkov, Richard A. Jonas, and the members of the CHSS*, Kansas City, Missouri; Toronto, ON, Canada; Cleveland, Ohio; Montreal, PQ, Canada; Boston, Massachusetts

Discussant: Frank L. Hanley, M.D.

OBJECTIVE: To determine outcomes and associated risk factors of different management strategies in the treatment of critical aortic stenosis (CAS) in neonates.

METHODS: Data regarding 285 cases (22 institutions) diagnosed within 30 days of birth from 1994 to 1999 were submitted to the CHSS Data Centre, and analyzed with parametric time-dependent event analysis in the hazard domain. Patients with aortic valve atresia, abnormal atrio-ventricular and ventriculo-arterial connections, or isolated aortic arch lesions were excluded.

RESULTS: Nineteen of 20 patients with no CAS-related intervention died (95%). Initial CAS-related procedure indicated a biventricular route (BVR) consisting of aortic valvotomy in 102 cases (31 died; 30%) or a "single ventricle" route (SVR)in 163 cases (59 died; 36%)(initial heart transplantation in 7, Norwood operation in 156). Anatomic characteristics completely discriminated BVR from SVR cases. Overall, survival in those having initial CAS-related intervention was 83% at 1 month after entry, 69% at 6 months, 66% at 1 year and 64% at 5 years. There was no significant difference in time-related death between BVR and SVR (survival at 5 years, 68% vs. 61%, respectively; p=0.36). Independent incremental risk factors for mortality included younger age at entry, non-Caucasian race, higher grade of MV hypoplasia, presence of LV endocardial thickening, and the presence of aortic coarctation (CoA). In SVR, incremental risk factors for mortality included non-Caucasian race, MV stenosis, and presence of VSD or CoA. In BVR, surgical valvotomy was associated with higher mortality than balloon valvotomy, but this was not significant after adjustment for anatomic complexity. Incremental risk factors for mortality in BVR included non-Caucasian race and higher grade of MV and LV hypoplasia. Freedom from CAS-related reintervention in BVR was 67% at 6 months and 44% after 5 years from initial valvotomy, with no difference between surgical vs. balloon valvotomy.

CONCLUSIONS: Neonatal CAS continues to be associated with a high mortality with both BVR and SVR. Better patient selection for both approaches may improve outcomes.

PRESENTATION OF SCIENTIFIC ACHIEVEMENT AWARD

Denton A. Cooley Houston, Texas

10:20 a.m. INTERMISSION - VISIT EXHIBITS

11:00 a.m. BASIC SCIENCE LECTURE:

Decoding the Human Genome

J. Craig Venter, Ph.D.

11:30 a.m. ADDRESS BY HONORED SPEAKER:

Effects of the Net Economy

James L. Barksdale

12:15 p.m. ADJOURN FOR LUNCH - VISIT EXHIBITS

12:30 p.m. CARDIOTHORACIC RESIDENTS' LUNCHEON

Metro Toronto Convention Centre, Summit Room

*By Invitation

TUESDAY AFTERNOON, MAY 2, 2000

1:45 p.m. SIMULTANEOUS SCIENTIFIC SESSION A - 2 ADULT CARDIAC SURGERY

Constitution Hall, Metro Toronto Convention Centre

Moderators: D. Craig Miller, M.D.

Timothy J. Gardner, M.D.

38. Early Discharge Following Coronary Artery Bypass Graft Surgery: Cost Savings or Cost Shifting?

Harold L. Lazar, Carmel A. Fitzgerald*, Tazeen Ahmad*, Yusheng Bao*, Oz M. Shapira*, and Richard J. Shemin, Boston, Massachusetts

Discussant: Richard D.Weisel, M.D.

OBJECTIVE: Changes in reimbursement policies and rising health care costs have resulted in shorter hospital length of stay(LOS) after Coronary Artery Bypass Graft(CABG) surgery. This study was undertaken to determine whether early discharge following CABG surgery resulted in cost savings or merely cost shifting by increasing the utilization of outpatient nursing and inpatient rehabilitation services.

METHODS: Patterns of discharge were analyzed in 330 patients undergoing CABG in 1990 (Group I) when there were no early extubation or fast track protocols and compared to 334 CABG patients in 1998(Group II) when these protocols were utilized.

RESULTS: Age, gender, angina class, ejection fraction, the incidence of diabetes, the number of vessels bypassed and the need for inotropic or mechanical support were similar between the groups. There was no difference in 30 day mortality, periop MI, infection, strokes or pneumonia. Group II patients spent less time on the ventilator(17.2 ±16.9 SD hours VS 10.219.5, P<0.001), had a lower incidence of reoperation for bleeding(0.5% VS 2.7%; P< 0.03), and had a shorter LOS (5.512.5 days vs 8.613.1; P<0.0001). However, fewer Group II patients were discharged home(55.3% VS 95.1%; P<0.000001). Furthermore more Group II patients went home with VNA services( 79.6% VS 7.5%; P<0.0001). A higher incidence of GROUP II patients (44.3% VS 2.4%; P<0.0001) were discharged to rehab facilities where their average LOS was 11.5±15.9 days. The need for readmission to acute care facilities was also increased in Group II patients( 5.3% VS 0.6%; P<0.0001).

CONCLUSIONS: Although early extubation and fast track protocols result in earlier discharge from acute care facilities, the anticipated savings may be offset by cost shifting due to increased (1) utilization of outpatient nursing services (2) discharges to rehab facilities and (3) hospital readmissions.

*By Invitation

39. Selection of a Cardiac Surgery Provider in the Managed Care Era

Winnie Yip*, David M. Shahian, Jerilynn Jacobson* and George A. Westcott*, Burlington and Boston, Massachusetts

Discussant: David F. Torchiana, M.D.

OBJECTIVE: Health planners predict that managed competition and the availability of outcome data should lead to more "rational" provider selection. Using a standard econometric model, we examine this hypothesis in the context of cardiac surgery, where there are profound implications for national health expenditures and quality.

METHODS: McFadden's conditional logit model was used to study the determinants of cardiac surgery provider selection among 6952 patients in the metropolitan Boston market (8 hospitals within a 15 mile radius) during fiscal year 1997. Hospital variables included beds, annual cardiac case volume, acuity-adjusted clinical and financial performance for the 3 preceding years (mean mortality, length of stay[LOS], charges, and cost), reputation markers (percent out-of-state referrals, cardiac residency program), and distance from the center of each hospital's zip code to the center of each patient's zip code. Patient variables included DRG, age, acuity level, and payer.

RESULTS: In all models, proximity of patient to hospital was the most stable and consistent predictor of choice (OR 0.89, p = 0.000). A cardiac surgery residency program significantly enhanced the probability of selection (OR 3.60, p = 0.000) as did percent out-of- state referrals (OR 1.06, p = 0.124). Higher adjusted mortality rates led to decreased probability of selection (OR 0.566, p = 0.113) but higher LOS was paradoxically associated with greater probability (OR 1.15, p = 0.193). Neither average hospital charges nor costs had any relationship to the probability of selection (OR 1.000, p = 0.989). Subgroup analysis by payer type revealed a striking increase in the preference for a teaching hospital among patients with commercial insurance (OR 19.08, p = 0.001). Non-Medicare managed care patients were the only subgroup in which higher mortality rate hospitals were more likely to be chosen (OR 3.34, p = 0.088).

CONCLUSIONS: Even within a competitive metropolitan market with advanced managed care penetration, the major observable determinants of cardiac surgery provider selection are hospital "reputation" and proximity to the patient's home, not objective clinical or financial performance.

*By Invitation

40. Reoperative Coronary Bypass Surgery: Effect of Patent Grafts on Perioperative Outcomes

Michael A. Borger*, Vivek Rao*, Richard D. Weisel, Alex Floh*, Gideon Cohen*, Christopher M. Feindel and Terrence M. Yau*, Toronto, ON, Canada

Discussant: Hendrick B. Earner, M.D.

OBJECTIVE: Patent grafts may increase the risk of reoperative coronary bypass surgery, an effect which may be mitigated by the use of retrograde cardioplegia. We attempted to determine the effects of patent grafts and retrograde cardioplegia on operative mortality.

METHODS: Systematic review of all redo coronary bypass patients (REDO, n = 744) at our institution from 1990-97. Independent predictors of operative mortality (OM) were determined with stepwise logistic regression analysis.

RESULTS: OM occurred in 42 patients (5.7%). Fifty percent of REDO patients had one or more patent grafts to the LAD, 33% to the RCA territory, and 27% to the circumflex territory. The previous LAD graft conduit was a saphenous vein in 82% and a LIMA in 18%. Patent grafts were injured in 14 patients (1.9%). Patent LAD grafts at the time of REDO did not result in a significant increase in the risk of OM (see Table), nor did patent grafts to the RCA or circumflex territory. Independent predictors of OM were age, NYHA class, LV grade, peripheral vascular disease, and failure to use retrograde cardioplegia (RETRO). RETRO was used in 40% of patients, and resulted in a significant decrease in OM (p = 0.02). Patients with stenosed LAD grafts seemed to receive the largest benefit from RETRO (OM 4% vs 10% without RETRO, p = 0.05).

CONCLUSIONS: We were unable to demonstrate an increased risk of operative mortality in redo patients with patent grafts. We strongly recommend the use of retrograde cardioplegia in reoperative coronary bypass surgery, particularly in patients with diseased LAD grafts.

Graft to LAD	Operative Mortality
None	10.7%
Stenosed	7.3%
Patent	4.7%
Occluded	3.8%

41. Long-Term Angiographic Follow-up of Complementary Saphenous Vein Grafting

Robert A. Dion*, David Glineur*, David Derouck*, Robert Verhelst*, Philippe Noirhomme*, Gebrine El Khoury* and Claude Hanet*, Brussels, Belgium

Discussant: Stephen E. Fremes, M.D.

OBJECTIVE: In order to achieve complete myocardial revascularisation, saphenous vein grafting (SVG) is still frequently used in addition to arterial grafting. We wanted to know the angiographic patency rates of complementary SVG after 10 years or more.

METHODS: Five hundred patients having received sequential internal thoracic artery grafting and complementary SVG between 1985 and 1991 were recently reviewed. Age averaged 61 years, 53 had a LVEF < 40%, 117 were operated in emergency, there were 35 reoperations. In total 2,156 distal anastomoses were constructed (4.3/patient), of whom 1,367 arterial (2.7/pt) and 789 venous (1.6/pt). Only 10.7% of the later were constructed on the LAD. The follow-up is 97.4% complete and averages 9.6 years. One hundred sixty-one patients consented to a late angiographic restudy after a mean interval of 7.4 y (max 12.2 y)

RESULTS: At 5 and 10 years, 94% and 77% of the patients remained free of cardiac events. Only 15 pts required an iterative revascularisation (CABG 4, PTCA 11), that is 0.3%/pt/year. Overall 428/448 arterial anastomoses (95.5%) and 153/211 venous anastomoses (72.5%) were patent, p <0.001. The sequential venous anastomoses remained strikingly more patent than the single grafts: 126/166 (76%) versus 27/45 (60%), p= 0.04. There was no difference in patency between the latero-lateral (diamond-shaped or not) and the termino-lateral (T or not) sequential anastomoses. There was no significant difference in patency between the anastomoses sequential or not directed to the LAD, Circumflex and right coronary artery areas: 16/19 (84.2%), 55/83 (66.3%) and 82/109 (75.2%), Pearson p = 0.2. Diabetes had no influence on patency rates neither overall (27/39, 69.2% versus 126/172, 73.2%: p = 0.8) nor for the sequential anastomoses (104/135, 77% versus 22/31, 71%: p = 0.5).

CONCLUSIONS: Complejmentary sequential venous grafting yields surprisingly high long term patency rates (76%). We could not find any influence of diabetes on the patency rates.

3:05 p.m. INTERMISSION - VISIT EXHIBITS

*By Invitation

3:25 p.m. SIMULTANEOUS SCIENTIFIC SESSION A - 2 ADULT CARDIAC SURGERY

Constitution Hall, Metro Toronto Convention Centre

Moderators: D. Craig Miller, M.D.

Timothy J. Gardner, M.D.

42. Myocardial Revascularization on the Beating Heart After Recent Onset of Acute Myocardial Infarction.

Giuseppe D'Ancona*, Hratch L. Karamanoukian*, Marco Ricci*, Reginald Abraham*, Jacob Bergsland* and Tomas A. Salerno, Buffalo, New York

Discussant: Gerald D. Buckberg, M.D.

OBJECTIVE: Coronary artery bypass grafting (CABG) after the recent onset of acute myocardial infarction (AMI), is associated with high morbidity and mortality. Revascularization without cardiopulmonary byapss (CPB), has been used to treat such patients (pts).

METHODS: From January 1995 to June 1999, 518 pts underwent CABG after recent AMI (1-20 days): CPB was used in 421 pts (Group A) and 97 pts (Group B)were operated without CPB. Preoperative risk factors were significantly (P<0.05) higher in Group B (redo, CHF, stroke, extensively calcified aorta,dialysis, evidence of left ventricular hypertrophy). Preoperative use of intra-aortic balloon pump (IABP) (5.2 vs. 2.4% P=NS) and emergent operations (5.2vs.2.6% P=NS) were similar in both groups. Mean number of grafts per pt was 3.46 in Group A vs. 1.82 in Group B (P=NS).

RESULTS: Crude mortality was 2.9% in Group A vs. 6.2% in Group B (P=NS). Major complications were comparable in the groups. Using univariate analysis mortality was found to be correlated to preopertive CHF, preoperative use of IABP, ventricular hypertrophy, advanced age, postoperative sepsis, and sternal infection. Multivariate analysis showed that only advanced age, postoperative sternal infection and sepsis, preoperative hemodynamic instability and evidence of left ventricular hypertrophy were positively related to death. Use or avoidance of CPB were not correlated to mortality when univariate or multivariate analysis was performed. Postoperative transmural AMI was positively related in univariate analysis to global ischemic time, preoperative HTN, female sex, use of warm cardioplegia and postoperative sepsis. Using logistic regerssion, global ischemic time was still correlated to postoperative trans-mural myocardial infarction together with preoperative HTN. The number of grafts was not correlated to postoperative AMI.

CONCLUSIONS: Multivariate analysis of a cohort of pts with recent AMI demonstrates that CABG can be performed with equal efficacy with oru without CPB. While CPB is not correlated to mortality, myocardial ischemic time (ie. cross-clamp time) is an independent risk factor for postoperative AMI.

*By Invitation

43. Bilateral Internal Mammary Artery Grafting: in Situ vs Y Graft. Long Term Clinical and Angiographic Results.

Antonio M. Calafiore, Marco Contini*, Giuseppe Vitolla*, Michele Di Mauro*, Valerio Mazzei*, Giovanni Teodori* and Gabriele Di Giammarco*, Chieti, Italy.

Discussant: Alfred J. Lector, M.D.

OBJECTIVE: To evaluate if the use of BIMA as in situ or Y graft provides the same long term results.

METHODS: From September 1991 to August 1999,1359 pts had BIMA in situ (n=1104, group A) or as Y (n=255, group B) graft.

RESULTS: Anastomoses/pt and BIMA anastomoses/pt were higher in group B (3.2±0.9 and 2.7±1.1) than in group A (2.8±0.8 and 2.3±0.8), p<0.001. Thirty day mortality was 1.9% in group A vs 3.1% in group B, p=ns. There was no difference in postoperative course. Eight years survival was 96.9+0.6 in group A vs 96.2±2.2 in group B, p=ns, and event free survival was 95.8±0.8 in group A vs 95.4±2.1 in group B, p=ns. Early angiographies were obtained in 281 pts (894 anastomoses, 828 distal and 66 proximal Y) 215 (591) in group A and 66 (303) in group B. Patency rate was 98.9% in group A and 96.7% in group B, p=ns. Late angiographies were obtained in 68 pts (18 in goup A and 50 in group B)at mean of 17.5±18.4 months: patency rate was 100% in group A and 98.9% in group B, p=ns, while grade A patency rate was 98.0% in group A and 98.4% in group B, p=ns. No Y anastomosis was occluded or stenosed.

CONCLUSIONS: The use of BIMA in situ or as a Y graft have similar survival, cardiac events incidence and angiographic patency in early and late phase.

*By Invitation

4:35 p.m. EXECUTIVE SESSION (Members Only)

Constitution Hall, Metro Toronto Convention Centre

6:15 p.m. Reception at Royal York Hotel

Followed By "The Lion King"

Princess of Wales Theatre

(Separate Subscription Required)

By Invitation

1:45 p.m. SIMULTANEOUS SCIENTIFIC SESSION B - 2

GENERAL THORACIC SURGERYRoom 205

Metro Toronto Convention Centre

Moderators: Mark J. Krasna, M.D.

David J. Sugarbaker, M.D.

45. Superficial Adenocarcinoma of the Esophagus

Thomas W. Rice, Malcolm M. DeCamp*, Gary W. Falk*, John R. Goldblum*, Adrian H. Ormsby*, David J. Adelstein*, Lisa A. Rybicki* and Eugene H. Blackstone, Cleveland, Ohio

Discussant: Nasser K. Altorki, M.D.

OBJECTIVE: Superficial adenocarcinoma of the esophagus (SAE), invading no deeper than the submucosa, was uncommon before the epidemic of Barrett's adenocarcinoma. SAE is now identified more frequently with regular endoscopic surveillance (ES); however, there is limited experience with treatment and outcome. The purpose of the study was to evaluate the results of surgical management of SAE and identify predictors of survival.

METHODS: Between 9/85 and 9/99, 111 patients (pts) underwent resection of SAE. 89% were men, median age was 64 years (yrs)(range 35-83). 50% were in ES programs. The table shows pathologic staging. Follow-up extended to 14 yrs, mean 47±40 months. Risk factors for mortality were identified by Cox multivariable regression.

RESULTS: There were 3 (2.7%) operative deaths. 5-yr survival decreased as depth of tumor invasion (T) increased: 96±4% Tis, 84±6% T1 intramucosal, and 47±12% T1 submucosal, P=.004. 5-yr survival was worse for N1 vs. No pts (17±16% vs. 82±5%, P<.001). 5-yr survival was worse in SAE discovered at first diagnostic esophagoscopy than in SAE discovered in ES programs (69±8% vs. 84±6%, P=.005). By multivariable analysis, N1 disease (P=.006), increasing T (P=.05), and older age (P=.007) decreased survival; participation in an ES program (P=.009) improved survival. Need for postoperative reinrubation led to worse early survival (47%±15 vs. 95%±2 at 1 yr, P=.02).

CONCLUSIONS: 1) Resection of SAE offers excellent survival with minimal operative mortality. 2) Survival is improved by ES and early resection before SAE invades the submucosa or metastasizes to regional lymph nodes. 3) Early survival may be improved by careful preoperative respiratory evaluation and aggressive perioperative respiratory care.

	Total n,(% total)	NO n,(% T subgroup)	N1 n,(% T subgroup)
Tis (high-grade dysplasia)	35 (32%)	35(100%)	0 (0%)
T1 intramucosal	47 (42%)	46 (98%)	1 (2%)
T1 submucosal	29 (26%)	23(79%)	6(21%)

*By Invitation

46. Histology and Stage Are Independent Prognostic Factors in Thymomas

Cameron D. Wright, Abeel A. Mangi*, John C. Wain, Dean M. Donahue*, James S. Allan*, Ashby C. Moncure, Earle W. Wilkins, Hermes C. Grille and Douglas J. Mathisen, Boston, Massachusetts

Discussant: Antoon E.M.R. Lerut, M.D.

OBJECTIVE(s): The Masaoka Staging system is currently used to stratify patients with thymomas. Histologic classification by the Muller-Hermelink Scheme has also been shown to correlate with prognosis. We reviewed patients with thymomas to evaluate the Masaoka staging system and histolgy as prognostic factors.

METHODS: Single institution restrospective review.

RESULTS: From 1972 to 1999, 155 patients underwent resection of a thy-moma. Overall 15 year survival was 55% whereas 15 year disease-specific survival (DSS) was 89%. Univariate analysis revealed that Masaoka stage (p<.0001), histology (p<.0001), and complete resection (p<.0001) predicted survival. Multivariate analysis revealed that Masaoka stage (p=.005) and histology (p=.02) independently predicted survival. There was no difference in disease-free survival (DPS) or DSS between Masaoka stage 1 or 2 or between 2a and 2b (p=ns). Classification of patients into two risk groups based on Masoka stage and histology clearly separated patients who had no relapses from those who did not (p=.0001) and was an independent predictor of survival (p=.002).

CONCLUSIONS: Histology (by the Muller-Hermelink system) is an independent predictor of survival in thymoma. The early stages of the Masaoka classification system are not distinct and do not accurately predict recurrences. A dichotomous classification system which takes into account both histology and stage better separates thymoma patients into clinically important prognostic groups and could help guide adjuvant treatment.

Risk of Recurrence Based on Stage and Histology

	LOW RISK			HIGH RISK
HISTOLOGY	Medullary	Cortical	WDTC	Cortical	WDTC
STAGE	1,2a,2b	1,2a		2b,3,4
RECURRENCE	0/88			15/67
DSS(15y)	98%			75%

*By Invitation

47. Neoadjuvant Chemotherapy Increases the Length of Stay and Health-Provider Effort in Patients Undergoing Pulmonary Resection for NSCLC

John R. Roberts*, Chad wick Eustis*, Elaine M. Eustis* and Walter Merrill, Nashville, Tennessee

Discussant: Keith S. Naunheim, M.D.

OBJECTIVE: Surgical effort has been the topic of HCFA and Medicare actions in recent months. However, little data about need for increasing surgical effort due to advances or changes in treatment exist. One such change is the use of neoadjuvant chemotherapy, which has become the standard for stage IIIA NSCLC in many institutions. Further, neoadjuvant therapy and may be used for earlier stages in the future. We have previously shown that neoadjuvant chemotherapy increases life-threatening complications in patients undergoing surgery and postulated that these patients would require greater surgical effort than other patients.

METHODS: All patients undergoing resection (lobectomy or greater) after neoadjuvant chemotherapy were compared to patients undergoing similar resections without preoperative chemotherapy. The resections were all done at a single institution in one year. Data collected were length of stay, ICU days, intubated days, chest tube duration, operative time, EBL, and health care provider visits. Two-tailed Student's t test was used to analyze differences in means and chi-square to determine differences in proportions. Differences <0.05 were considered significant.

RESULTS: Thirty-four patients underwent resection after neoadjuvant chemotherapy and 67 patients were resected without preoperative therapy. No differences between the groups in age, pulmonary function, or comorbid diseases were found. The patients receiving chemotherapy did have a more advanced stage (2.52 versus 1.55, p<0.0001). There was no hospital mortality. Patients receiving preoperative chemotherapy had a greater length of stay (13.9 vs. 8.0 days, p=0.032), greater blood loss (462 cc vs 304 cc, p=0.03), and required twice as many health provider visits (66.1 vs 33.4, p=0.03)

CONCLUSIONS: Neoadjuvant carboplatin and taxol increased the length of stay, EBL, and number of physician visits in this cohort of patients compared to a similar cohort undergoing surgery in the same institution. These data demonstrate that neoadjuvant chemotherapy increases the physician effort necessary to care for these patients.

2:45 p.m. INTERMISSION - VISIT EXHIBITS

*By Invitation

3:30 p.m. SIMULTANEOUS SCIENTIFIC SESSION B - 2

GENERAL THORACIC SURGERY Room 205

Metro Toronto Convention Centre

Moderators: Mark J.Krasrw,M.D.

David J. Sugarbaker, M.D.

48. Outcome of Lung Volume Reduction Surgery in Emphysema Patients Eligible for Lung Transplant

Bryan F. Meyers*, Stephen S. Lefrak*, Mary S. Pohl*, Tracey J. Guthrie*, Roger D. Yusen*, G. Alexander Patterson and Joel D. Cooper, St. Louis, Missouri

Discussant: Douglas E. Wood, M.D.

OBJECTIVE: Between March 1993 and May 1998, we performed 200 consecutive bilateral lung volume reduction (LVRS) operations for patients with emphysema. Ninety-nine of these patients were considered eligible for either LVRS or lung transplant (TX) based on age, impairment, and absence of contraindications. The clinical outcomes of these 99 patients were reviewed to assess the consequences of LVRS on patients eligible for lung transplant.

METHODS: A retrospective chart review was performed using a prospectively assembled computer database.

RESULTS: The 61 men and 38 women had a mean age of 55 ± 7 years at the time of LVRS. Mean values for first second expired volume, total lung capacity and residual volume were 24 ± 7, 141 ± 19 and 294 ± 54 percent predicted, respectively. These values are identical to those observed in our overall LVRS experience. There were 4 perioperative deaths and 14 late deaths. Two-year and 5-year survival after LVRS were 92% and 77%. The 31 patients who have been listed for TX after LVRS include 13 who have been transplanted, 14 who remain on the list, and 4 who have been removed from the list. All 13 transplanted patients survived TX and one has subsequently died of chronic rejection. Twelve surviving recipients have a median post-transplant follow-up of 1.3 years. The mean age of the TX recipients was 57.5 ± 5.1 years with TX occurring 3.6 ± 1.0 years after LVRS. The fourteen patients still on the TX waiting list have a mean interval since LVRS of 4.6 ± 1.0 years. Sixteen of the 99 patients underwent lower lobe LVRS and 10 of these patients have either been transplanted (6) or listed (4). No significant differences were found between patients listed for TX and patients not listed when compared according to age, lung function, or response to LVRS.

CONCLUSIONS: Lung transplant following LVRS is feasible and associated with no apparent increased mortality. Patients undergoing LVRS for lower lobe disease are more likely to progress to transplant.

*By Invitation

49. A Novel Approach Using Magnetic Resonance Technique for the Detection of Lung Allograft Rejection

Shinichi Kanno*, Paul C. Lee*, Stephen Dodd*, Mangay Williams*, Timothy R. Billiar*, Bartley P. Griffith, Chien Ho*, Pittsburgh, Pennsylvania

Discussant: Steven J. Mentzer, M.D.

OBJECTIVE(s): Although various techniques have been explored for the detection and quantification of allograft rejection, a practical and reliable method that is non-invasive is still elusive.

METHODS: For our magnetic resonance (MR) experiments, we have developed a new rat model of heterotopic lung transplantation to the inguinal region. Allogeneic transplants (DA→BN) were performed with and without cyclosporin-A (CsA) treatment, with syngeneic transplants (BN→BN) serving as controls (n=6 per group). MR images were obtained with a gradient echo method before and after injection of ultra-small superparamagnetic iron oxide (USPIO).

RESULTS: At day 5, allogeneic transplants without CsA treatment developed a grade 4 rejection pathologically. A significantly lower MR signal was seen 24 hours after USPIO injection (346±7.6 vs 839±43.4, arbitrary unit, p<0.05). Syngeneic transplants showed no evidence of rejection pathologically and no differences in MR signal between injections (863±18.8 vs 880±22.5). Allotransplants treated with CsA showed a grade 2 rejection pathologically. The change in MR signals in that group was small, but significant enough to show a decrease in signal intensity after injection (646±10.5 vs 889123.5, p<0.05). Immunohistochemistry and iron staining in the allografts revealed that USPIO was taken up by the infiltrating macrophages that accumulated at the rejecting site.

CONCLUSIONS: We demonstrated a novel approach for detection of acute lung rejection with USPIO injection. This method might have tremendous clinical application.

*By Invitation

50. Laryngotracheal Resection and Reconstruction for Postintubation Tracheal Stenosis Extending to the Subglottic Region

Paolo Macchiarini *, Jean-Philippe Verroye *, Alain Chapelier *, Elie Fadel * and Philippe Dartevelle, Hannover, Germany; Paris, France

Discussant: F. Griffith Pearson, M.D.

OBJECTIVE: Analyze the characteristics and results of laryngotracheal resection and reconstruction for postintubation tracheal stenosis extending to the subglottic region.

METHODS: Fourty-rwo patients (31 males and 11 females, mean age 41±18 years) underwent resection of the anterior cricoid cartilage and primary thyrotracheal reconstruction for subglotic stenosis. Five of them had also a tracheo-esophageal fistula, repaired simultaneously via a primary two-layers esophageal closure. Twenty-patients (52%) were referred to us after initial unsuccessful endoscopic (n=22) or surgical (n=5) management. The stenosis appeared 30±43 days from endotracheal intubation (n=19) or tra-cheostomy (n=23). There were 24 cuff lesions, 7 stomal lesions, and 11 at both levels; all but one stenosis were circumferential. The upper limit of the stenosis lied 1.9±.7 cm below the vocal cords. Stenoses measured 2.8±.9 cm in length and the esophageal defects 2.5±1 cm. The subglottic diameter was reduced by 60% in 36 or 86% of patients. All but one operations were performed through a cervical incision only. The length of resection ranged from 2 cm to 6.5 cm (mean 4.6±1). Eighteen thyrohyoid and 4 supralaryngeal releases were employed to reduce anastomotic tension.

RESULTS: Four patients (9%) required post-operative tracheostomy, and 38 (91%) were extubated within 24 hours. Early complications occurred in 11 patients (27%) and were most frequent in patients requiring laryngeal release and extended resections; one patient died (2%). Among the remaining 41 patients, 39 or 95% had excellent or good anatomical and functional long-term results. Two failures required definitive tracheostomies.

CONCLUSIONS: Resection of the anterior cricoid cartilage and primary thyrotracheal reconstruction is the best treatment for post-intubation sub-glottic stenosis.

*By Invitation

51. Timed Barium Esophagram: a Simple Physiologic Assessment for Achalasia

Srodjan Kostic*, Thomas W. Rice, Joel E. Richter*, Mark E. Baker*, Malcolm M. DeCamp*, Lisa A. Rybicki* and Eugene H. Blackstone, Cleveland, Ohio

Discussant: Thomas R. J. Todd, M.D.

OBJECTIVE: The outcome of achalasia therapy is difficult to measure because repeated physiologic study is impractical and symptom interpretation is subjective. In contrast, timed barium esophagram (TBE) is simple, easily performed, inexpensive, quantitative, repeatable, and comfortable for the patient. The purposes of this study were 1) to evaluate the use of TBE in appraising the outcome of myotomy and 2) to determine the cause of symptoms and their relief by myotomy.

METHODS: 52 patients (pts) ingested 250 ml low-density barium and had upright spot films at 1, 2, and 5 minutes preoperatively (preop) and at 8 weeks (median) after myotomy (postop). Height and width of the barium column and their change over time were measured and related by multi-variable analyses to symptoms of regurgitation, dysphagia and chest pain. Symptoms were scored as 0 (none) to 5 (continuous).

RESULTS: At 1,2, and 5 minutes preop, mean barium column height was 18,16, and 15 cm, and width 5.7,5.3, and 5.0 cm. Surgery reduced these to 7.7, 6.6, and 4.5 cm, and 3.4, 3.1, and 2.6 cm, respectively (P<.001). The preop degree of regurgitation was related directly to height of the barium column at 1 minute (P=.003).Mean height was 23 cm for grade 4-5 regurgitation and 14 cm for grade 0-1. Degree of dysphagia was related directly to change in width from 1 to 5 minutes (P=.06).Mean change was -0.4 cm for grade 4-5 dysphagia (P=.06)and -1.0 cm from grade 0-1. Chest pain was related inversely to width at 1 minute (P=.06). Width was 5.0 cm for pts with any grade chest pain and 6.1 cm for those without. Surgery relieved symptoms in the majority of pts (grade 4-5 regurgitation from 45% to 5%, grade 4-5 dysphagia from 86% to 7%, and any grade chest pain from 52% to 12%, P<.001). Residual symptoms were unrelated to TBE measurements.

CONCLUSIONS: 1) TBE gives objective confirmation of successful myotomy. 2) TBE reveals that regurgitation and its relief are related to the height of the barium column, dysphagia to the rate of esophageal emptying, and chest pain to the less dilated non-myotomized esophagus. TBE is a simple measure of esophageal emptying that elucidates the mechanism of symptoms and their relief by myotomy.

*By Invitation

4:35 p.m. EXECUTIVE SESSION (Members Only)

Constitution Hall, Metro Toronto Convention Centre

6:15 p.m. Reception at Royal York Hotel

Followed By "The Lion King"

Princess of Wales Theatre

(Separate Subscription Required)

1:45 p.m. SIMULTANEOUS SCIENTIFIC SESSION C - 2

CONGENITAL HEART DISEASE Room 201

Metro Toronto Convention Centre

Moderators: John J. Lamberti, M.D.

William G. Williams, M.D.

52. Modifications to the Cavopulmonary Anastomosis Do Not Eliminate Sinus Node Dysfunction

Mitchell I. Cohen*, Nancy D. Bridges*, J. W. Gaynor*, Timothy M. Hoffman*, Gil Wernovsky*, Victoria L. Vetter*, Thomas L. Spray and Larry A. Rhodes*, Philadelphia, Pennsylvania

Discussant: Peter Manning, M.D.

OBJECTIVE: Sinus node dysfunction (SND) occurs frequently after the Fontan (F) operation and can have deleterious effects on F physiology. The objective of this study was to determine whether modifications of the cavopulmonary anastomosis (CPA) which avoid surgery near the sinus node result in a lower incidence of SND.

METHODS: Since 1996, a prospective cohort study has been conducted evaluating the incidence of SND in all patients (pts) staged with either an initial hemi-Fontan (HF) or bidirectional Glenn (BDG) and a subsequent lateral tunnel (LT) or extracardiac conduit (EC). Only pts with normal sinus node function prior to the HF or BDG were included. SND was defined by a heart rate ≥ 2 S.D. below age adjusted norms, or predominant junctional rhythm, or a sinus pause ≥ 3 sec. as determined by resting ecg and/or 24° Holter monitor at hospital discharge (d/c). The primary outcome evaluated was the difference in the incidence of SND in pts with either HF/LT or BDG/EC. Other perioperative characteristics which might lead to SND were evaluated.

RESULTS: As of 9/99, 74 pts have entered the study: 46 had a HF (mean age 6.8±2 mos) and 28 a BDG (mean age 7.3±3 mos). All 28 BDG pts and 2 additional HF had a subsequent EC F (mean age 24±5 mos). The remaining 44 HF pts had a LT F (mean age 23±6 mos). Diagnoses were, hypoplas-tic left heart (29), single LV (18), single RV (13), heterotaxy (7), and other (7). Among the 74 pts, SND was present in 9 (12%) after the F. In those with a HF/LT the incidence was 14% (n=6) and in BDG/EC pts the incidence was 10% (n=3) (.95 CI .05-.27 vs .02-.28; risk ratio 1.3; p=NS). No diagnostic or intraoperative variables were associated with SND. There were no deaths in the BDG/EC group and 1 death in the HF/LT group. 2 HF/LT pts had pacemakers: 1 for heart block (normal atrial rate)and 1 for prolonged sinus pauses.

CONCLUSIONS: In this cohort study, avoidance of surgery near the sinus node had no discernible effect on the development of SND. Thus concerns about SND should not override pt anatomy or surgeon preference as determinants of which CPA to perform. Longer follow-up is needed to determine if either staging strategy will reduce the long-term incidence of SND and pacemaker implantation.

*By Invitation

53. Pulmonary Microvessel Density Is a Marker of Angiogenesis in Children After Cavopulmonary Shunt

Sandra L. Starnes*, Brian W. Duncan*, James M. Kneebone*, Shawn States*, Geoffrey L. Rosenthal* and Flavian M. Lupinetti*, Seattle, Washington

Discussant: Richard A. Jonas, M.D.

OBJECTIVE: Pulmonary arteriovenous malformations are a frequent cause of progressive cyanosis in children after cavopulmonary anastomosis and may represent a form of abnormal angiogenesis. Microvessel density has been used as a marker of angiogenesis in tumor studies. We determined the microvessel density in the lungs of children after cavopulmonary anastomosis with and without clinical evidence of pulmonary arteriovenous malformations to determine if they had increased numbers of blood vessels consistent with ongoing angiogenesis.

METHODS: Lung biopsy specimens were obtained from six children following cavopulmonary shunt and four age-matched controls. Of the six children following cavopulmonary anastomosis, two had angiographically documented pulmonary arteriovenous malformations while the other four children had no clinical or angiographic evidence of pulmonary arteriovenous malformations. Microvessels staining positive for a primary antibody to von Willebrand factor were counted in ten high-power fields (200X) per patient.

RESULTS: The mean (±standard error) microvessel density in all patients following cavopulmonary anastomosis was 31.3±7.4 versus 9.3±5.7 in controls, p =0.02. There was no difference in microvessel density in children with and without pulmonary arteriovenous malformations following cavopulmonary anastomosis (33.1±10.6 versus 30.4±8.6, p =0.9).

CONCLUSIONS: Following cavopulmonary anastomosis, children have greatly increased numbers of pulmonary microvessels regardless of whether they have clinically or angiographically significant pulmonary arteriovenous malformations. This supports the evidence of a constant angiogenic stimulus in the lungs of children after cavopulmonary anastomosis.

54. Long-Term Results of the Lateral Tunnel Fontan Operation

Christof Stamm*, Ingeborg Friehs*, John E. Mayer, David Zurakowski*, John K. Triedman*, Edward P. Walsh*, Richard A. Jonas and Pedro J. del Nido, Boston, Massachusetts

Discussant: Gordon K. Danielson, M.D.

OBJECTIVE: Construction of a total cavopulmonary anastomosis using an intra-atrial lateral tunnel is known to yield good early and mid-term results. Given the current controversy regarding indications for a total extracardiac Fontan procedure, we reviewed the long-term outcome after a lateral tunnel Fontan operation.

METHODS: Between 10/87 and 12/91, 220 patients (age: 11 months to 32 years, median = 3.9 ± 0.6 years) underwent a fenestrated or non-fenestrated lateral tunnel Fontan procedure at our institution. Diagnoses included single left ventricle with normally related (n=72) or transposed (n=81) great arteries, single right ventricle (n=28), heterotaxy (n=22), hy-poplastic left heart (n=14), and others (n=3). Current follow-up information was available for 179 patients (mean follow-up 10.2 ± 0.6 years). Risk factor analysis included patient- and procedure-related variables with death, failure (death, takedown, or transplantation), and brady- or tachyarrhythmia as outcome parameters.

RESULTS: There were 10 early deaths, 2 late deaths, 4 take-down operations, and 2 heart transplantations. Kaplan-Meier estimated survival was 92.7% (70% confidence interval = 91-95%) at 5 and 10 years, freedom from failure was 89% (86-91%) at 5 years and 88% (85-90%) at 10 years. Freedom from bradyarrhythmia was 93% (91-95%) at 5 years and 89% (87-92%) at 10 years, freedom from atrial tachyarryhthmia 98% (97-99%) at 5 years and 94% (92-96%) at 10 years. One patient developed protein losing enteropathy. Risk factors for development of atrial tachyarrhythmia were heterotaxy syndrome (odds ratio = 14.1, P = 0.002) and single morphologic right ventricle (OR = 7.6, P = 0.01). None of the patient-related variables significantly influenced survival.

CONCLUSIONS: The lateral tunnel Fontan operation results in superior long-term survival irrespective of the underlying anatomic diagnosis. The incidence of atrial arrhythmia appears to depend on ventricular morphology. The excellent long-term outcome after an intracardiac lateral tunnel Fontan procedure should serve as a basis for comparison with other surgical alternatives.

*By Invitation

55. Modified Norwood Procedure Using a High Flow Cardiopulmonary Bypass Strategy Results in Low Mortality Without Late Arch Obstruction.

Nancy C. Poirier*, Jonathan J. Drummond-webb*, Michi Imamura*, Alexander M. Harrison*, Roger B. B. Mee, Cleveland, Ohio

Discussant: Edward L. Bove, M.D.

OBJECTIVE: We reviewed our results of a modified stage 1 Norwood repair (mNr) using only autologous tissue. It is performed with high flow Cardiopulmonary bypass (CPB)followed by aggressive postoperative va-sodilation (both attained using phenoxybenzamine) and a normocapneic ventilatory strategy.

METHODS: Between 1993 and 1999, 59 patients aged 1 to 353 days (median 4 days) and weighing 1.7 to 6.8 kg (median 3.2 kg)underwent a mNr. The procedure consists of excising ductal tissue and augmenting the arch by the combined anastomosis of the main pulmonary artery and descending aorta. Ascending aortic diameter ranged from 1.5 to 8 mm (median 3). The modified Blalock-Taussig shunt was 3 mm in 21 patients (36%) and 3.5 mm or larger in 38 patients (64%).

RESULTS: Circulatory arrest (CA) and CPB times ranged from 15 to 64 min (median 37) and 44 to 144 min (median 88) respectively. Postoperative survival was 83% (49/59). At univariate analysis, early mortality was associated with ascending aortic diameter of less than 2.5 mm (p=0.015). Weight, age at operation, associated procedures, CAand CPB times, diagnosis (HLHS vs variant), shunt size, and date of the procedure did not influence the early or late operative survival. During a median follow-up period of 39 months (range=l-63), 30 patients underwent bidirectional cavopulmonary shunts (BCPS; 61%), 6 Fontan (12%) and 1 cardiac transplantation following a failed BCPS. Four patients died for an overall survival of 76%. Neo-aortic or arch obstruction was corrected in 3 patients (5%).

CONCLUSIONS: Our results of mNr using this perioperative strategy are acceptable with a low incidence of neo-aortic and arch obstruction. Patients with small ascending aorta diameters are a high-risk group and perhaps candidates for alternative approaches.

2:45 p.m. INTERMISSION - VISIT EXHIBITS

*By Invitation

3:25 p.m. SIMULTANEOUS SCIENTIFIC SESSION C - 2

CONGENITAL HEART DISEASE

Metro Toronto Convention Centre

Moderators: John J. Lamberti, M.D.

William G. Williams, M.D.

56. Long-Term Follow-up After Early Repair of Tetralogy of Fallot: the Impact of a Transannular Patch.

Emile A. Bacha*, Lars Erickson*, John E. Mayer, Pedro J. del Nido, Judy Huing*, Peter Lang*, Richard A. Jonas, Boston, Massachusetts

Discussant: Roger B. B. Mee, M.D.

OBJECTIVE: Early primary repair of Tetralogy of Fallot (TOP) has been routinely performed at this institution since 1972. This study examines the impact of a transannular patch (TAP) on mortality, need for reoperation, right heart failure, medication, arrhythmias, and reproductive health in a cohort of patients 20 or more years following this procedure.

METHODS: Sixty patients aged less than 2 years underwent repair of TOP between 1/1972 and 12/1977 (median age 8.1 months). Follow-up data were obtained on 46 (85%) of the 52 patients who survived the operation (median follow-up 17.5 years). A TAP was used in 32 survivors (70%). Data were compared between the TAP-group and non-TAP group.

RESULTS: No late deaths were identified. Nine patients (17 %) required a reoperation: 6 (5 with non-TAP) underwent relief of right ventricular outflow tract obstruction. One patient each required a pacemaker, an AICD, and a pulmonary valve insertion for right ventricular dilation (all with TAP). Forty-two patients (91%) were in NYHA class I and 4 in class II. Two out of 16 females, both with TAP, carried pregnancies to term. Freedom from reoperation of any type was 94 vs. 88% at 10 years and 91 vs. 80% at 18 years for TAP vs. non-TAP (p=0.11). No independent predictors of long-term survival could be identified by multivariate analysis.

CONCLUSIONS: Pulmonary valve insertion is rarely indicated after early primary repair of TOP with a TAP. There is equal long-term survival and freedom from reoperation as compared to non-TAP.

*By Invitation

57. Coronary Arterial Size Late After Atrial Switch Operation in Patients with Transposition of Great Arteries: Implications for Arterial Switch Operation.

Zahid Amin*, Phillip Moore*, Doff B. McElhinney*, Vadiyala M. Reddy and Frank L. Hanky, Augusta, Georgia; San Francisco, California; Philadelphia, Pennsylvania

Discussant: Roger B. B. Mee, M.D.

OBJECTIVE: Coronary flow reserve in the hypertrophied ventricle is reduced. One contributing factor may be the size of proximal coronary arteries. In patients who have undergone an atrial switch procedure for transposition of great arteries, the left coronary artery (LCA) supplies the pulmonary ventricle and may be small. We hypothesized that the dimensions of the coronary arteries may be related to symptomatic status in these patients, and that a small LCA may relate to failure of attempted pulmonary artery banding (PAB) and conversion to arterial switch, possibly by affecting blood supply to the interventricular septum.

METHODS: Left and right coronary arteries (RCA) were measured in 10 asymptomatic patients after atrial switch procedure, 9 symptomatic patients after atrial switch procedure, and 10 patients with normal hearts. The size of the coronary arteries was indexed to the body surface area.

RESULTS: Both the absolute and indexed diameters of the LCA were significantly smaller than those of the RCA in symptomatic patients after atrial switch (indexed: 2.3±0.5 vs 3.3±0.6 mm, p<0.001), whereas there was no difference in asymptomatic patients (2.2±0.4 vs 2.4±0.4, p=0.47). Similarly, the RCAtLCA diameter ratio was significantly larger in symptomatic than asymptomatic patients (1.5±0.3 vs 1.1+0.3, p=0.007). There was no difference in coronary artery size between normal patients and asymptomatic patients with TGA after atrial switch procedure.

CONCLUSIONS: Differences in size of the proximal coronary arteries may be related to symptomatic status in patients with transposition of great arteries who have undergone atrial switch procedure, possibly by affecting blood supply to the interventricular septum.This would predispose to more severe tricuspid regurgitation and failure. When PAB and subsequent arterial switch are considered for patients with atrial switch and a failing systemic right ventricle, size of the LCA may be an important factor to consider, and should be evaluated with preoperative imaging studies.

*By Invitation

58. Ross Procedure in Children : a Word of Caution.

Antonio Laudito*, V. Mohan Reddy*, Michael M. Brook*, Marc S. Bleiweis*, Lenardo D. Thompson* and Frank L. Hartley, San Francisco, California

Discussant: Ronald C. Elkins, M.D.

OBJECTIVE: Aortic valve disease in children is often a difficult and complex problem with controversial management strategies. The Ross and Ross-Konno procedures have become the primary choice for aortic valve replacement (AVR) because of optimal hemodynamic performance, no need of anticoagulation and growth potential. However, there is continuing concern regarding the longevity of the pulmonary autograft especially in patients with primary aortic insufficiency (AI).

METHODS: Between June 1993 and September 1999, 77 Ross and Ross-Konno procedures were performed at our institution: 10 patients were infants (borderline hypoplastic left heart patients were excluded), 61 were children, 6 were adults. Preoperative, postoperative, and follow-up clinical, echocardiographic, and hemodynamic data were reviewed. Statistical analysis was performed to identify the risk factors for deteriorating autograft function.

RESULTS: Primary AI was an indication for Ross procedure in 23 patients and mixed lesion with predominant AI in 20 patients. In the rest aortic stenosis (AS) was the primary or predominant lesion. There were no early or late deaths. Prosthetic AVR was necessary in the follow-up period in 5 children for severe AI: 3 of them had AI as primary lesion, and 2 had mixed lesion with predominant AI in one. At follow-up 4 children have moderate AI: 3 had predominant AI as indication for Ross procedure and 1 had AS as primary lesion. Although it did not reach statistical significance (X²test, p= 0.077),there is a trend of deteriorating autograft function in patients with primary or predominant AI. In addition other reinterventions included right ventricle to pulmonary artery conduit replacement (n=7, including 5 pts during AVR) and internal mammary artery coronary bypass grafting (n=2).

CONCLUSIONS: Ross and Ross-Konno procedures have altered the prognosis of infants and young children with severe and complex aortic valve disease. However, Ross procedure in older children especially with primary or predominant AI should be performed with caution. Further follow-up to delineate the risk factors for autograft dysfunction in children is necessary to better define the indications for Ross procedure

*By Invitation

4:35p.m. EXECUTIVE SESSION (Members Only)

Constitution Hall, Metro Toronto Convention Centre

6:15 p.m. Reception at Royal York Hotel

Followed By "The Lion King"

Princess of Wales Theatre

(Separate Subscription Required)