79TH ANNUAL MEETING

Ernest N. Morial Convention Center

New Orleans, Louisiana - April 18-21, 1999

PROGRAM

MONDAY, APRIL 19, 1999

8:00 a.m. BUSINESS SESSION (Limited to Members)

8:15 a.m. SCIENTIFIC SESSION

Ballroom, Ernest N. Morial Convention Center

Moderators: Lawrence H. Cohn, M.D.

Tirone E. David, M.D.

1. Transmyocardial Revascularization Combined with Coronary Artery Bypass Grafting Versus Bypass Grafting Alone: A Prospective, Randomized Multi-Center Trial

Keith B. Allen*, Anthony J .Delrossi, Fidel Realyvasquez*, Edward A. Lefrak, Robert D. Dowling*, Thomas Pfeffer, Tommy Fudge, Mark Mostovych, Douglas Schuch and Szaboic Szentpetery, Carl J .Shaar*, Indianapolis, IN, Camden, Redding, CA, Falls Church, VA, Louisville, KY, Los Angeles ,CA, Houma, LA, Jacksonville, FL, Sacramento, CA, Norfolk, VA

Discussant: O. Howard Frazier, M.D., Houston, TX

OBJECTIVE: To assess the safety and efficacy of transmyocardial revascularization (TMR) combined with coronary artery bypass grafting (C ABG) in patients not amenable to complete revascularization by C ABG alone.

METHODS: Two hundred sixty-six patients at 24 centers whose standard of care was CABG, but who had one or more ischemic areas not amenable to bypass grafting were prospectively randomized. Group A (n=133) received CABG of suitable vessels plus TMR to areas not graftable. Group B (n=133) received CABG alone with non-graftable ischemic areas left unrevascularized. Patient demographics including preoperative ejection fraction (EF), distribution and number of bypasses, number of endarterectomies, cardiopulmonary bypass time and predicted operative mortality of 8.6% and 7.7% (Parsonnet analysis) were similar for both groups.

RESULTS: Group A (CABG + TMR) operative mortality was 1.5% (2/133) and significantly (p=0.02) lower than Group B (CABG alone) at 7.5% (10/133). Group A operative mortality was significantly lower than predicted (p<0.0001). Major 30-day cardiac events (death or myocardial infarction) for Groups A and B were 3% (4/133) and 9% (12/133), (p=0.05). Multivariable analysis of treatment arm (CABG + TMR vs CABG alone), EF, age, prior CABG, diabetes and gender determined CABG alone as the sole predictor of operative mortality (odds ratio: 5.45, 95% CI 1.3-26.4; p=0.03). At three-months, mortality remained significantly lower for Group A [3% (3/117)] than for Group B [10% (11/115)] with 88% and 86% follow-up, (p=0.03).

CONCLUSIONS: This prospective, randomized multi-center trial demonstrates that TMR combined with CABG is safe and reduces operative mortality in patients not amenable to complete revascularization by CABG alone.

*By Invitation

2. Eliminating the Cervical Esophagogastric Anastomotic Leak with a Side-to-Side Stapled Anastomosis

Mark B. Orringer and Mark D. Iannettoni*, Ann Arbor, Michigan

Discussant: Victor F. Trastek, M.D., Rochester, MN

While the acute postoperative complications associated with a cervical esophagogastric anastomosis (CEGA) are less than those with an intrathoracic esophageal anastomosis, the long-term sequelae of a CEGA leak have not proven to be as minor as initially reported. As many as 50% of CEGA leaks result in an anastomotic stricture, and the subsequent need for chronic dilatations negates the merits of an operation intended to restore comfortable swallowing.

OBJECTIVE: This study was undertaken to determine if construction of a side-to-side stapled CEGA after transhiatal esophagectomy (THE) could reliably eliminate the majority of anastomotic leaks.

METHODS: During the past 18 months, in 98 consecutive patients undergoing THE, a functional side-to-side CEGA was constructed with the endo-GIA stapler applied directly through the cervical wound. This side-to-side stapled anastomosis (SSSA) has 3 rows of staples. Early postoperative anastomotic morbidity, subsequent need for anastomotic dilatations, and patient satisfaction with swallowing were evaluated.

RESULTS: Prior to the use of the SSSA, regardless of the manual suture technique used to construct the CEGA, the anastomotic leak rate in over 1,000 patients undergoing THE varied from 10-15%. Among the 98 patients undergoing THE and a SSSA, there were 2 anastomotic leaks (2%). This dramatically lowered leak-rate has contributed to reduction in the currernt average length of stay after a THE to 7 days and has provided more comfortable swallowing, less need for subsequent esophageal dilatations and greater patient satisfaction.

CONCLUSIONS: Construction of the CEGA using a SSSA virtually eliminates anastomotic leaks and reduces the long-term morbidity of anastomotic stricture. The SSSA is a major technical advance in the progression of refinements of THE and a CEGA.

*By Invitation

3. Effect of Repair Strategy on Cost and Outcome for Infants with Hypoplastic Left Heart Syndrome: Palliation versus Heart Transplant

^†Ko Bando , Debby Murphy*, Theresa Flaspohler*, Mark W. Turrentine*, Thomas G. Sharp*, Randall L. Caldwell*, Robert K. Darragh*, Thomas X. Aufiero* and John W. Brown, Indianapolis, Indiana and Osaka, Japan

Discussant: Leonard L. Bailey, M.D., Loma Linda, CA

OBJECTIVE: Treatment of hypoplastc left heart syndrome (HLHS) remains expensive and controversial. Since 1989 we have performed either staged Norwood operations (Nw) or heart transplantation (Tx) for HLHS. This study sought to assess the effect of repair strategy on cost and outcome for HLHS.

METHODS: Preop, operative and follow-up costs (adjusted to 1998 US dollars) were obtained for 90 patients (pts) with HLHS from January 1989 to June 1998. Records were reviewed to determine late outcomes. Functional status was assessed by questionnaire.

RESULTS: Three groups were studied. Tx group (n=18): 12 before 1993. Nw Group1 (n=24): Before 1993, Nw was performed on 24 pts (6 cross-over from Tx). Nw Group2 (n=48): Since 1994, 48 pts had Nw (4 cross-over from Tx).

	Tx Group	Nw Group 1	Nw Group 2	P value
Preop hospital days	27.4±31.6	6.9±4.7	3.2±.8.5	†p<0.001
Pre-op cost *	94.6±72.1	32.0±14.7	10.4±8.0	†p<0.001
5yr total cost*	364.1±149.8Œ	191.0±39.1'	138.3±24.1'	†p<0.001
5yr actuarial survival (%)	72.7±0.2	37.5±0.2‡	71.1±0.2	‡p=0.002
NYHA class (mean)	1.2	1.2	1.2	NS

* in $1,000. Œ includes Tx and follow up. ´ includes stages 1, 2, 3 and follow-up.

†ANOVA, ‡Nw Group1 vs Tx or Nw Group 2 by log rank

CONCLUSIONS: Both Tx and Nw offer good late functional outcomes, but the total costs for Tx exceed Nw significantly. This plus the fact that late survival has significantly improved for Nw since 1994, make it the procedure of choice at our center.

†1992-92 AATS Graham fellow

*By Invitation

4. Glucose-Insulin-Potassium Solutions Improve Clinical Outcomes in Diabetic Patients Undergoing Coronary Bypass Grafting

Harold L. Lazar, Stuart Chipkin*, George Philippides*, Yusheng Bao* and Carl Apstein*, Boston, Massachusetts

Discussant: Richard D. Weisel, M.D., FRCS, Toronto, Ontario, Canada

OBJECTIVE: Patients with diabetes mellitus have significantly higher morbidity following Coronary Artery Bypass Grafting (CABG). Our previous studies have shown that Glucose-Insulin-Potassium (GIK) solutions improve clinical outcomes in non-diabetic patients following urgent CABG. This study was therefore undertaken to determine whether substrate enhancement with GIK would improve myocardial performance and limit morbidity following CABG in diabetic patients.

METHODS: Forty consecutive CABG patients with medically treated diabetes mellitus (tablets or insulin) were prospectively randomized into a GIK group: N=20; (500 ml D5W+80 units regular insulin + 40 mEq KCL at 30 ml/ hr) and a NO-GIK group: N=20; (D5W at 30 ml/hr). GIK was instituted upon anesthetic induction and continued for 12 hours post-op.

RESULTS: The two groups did not differ statistically in gender, number of vessels bypassed, angina class, ejection fraction, pre-op cardiac index, preop serum glucose, crossclamp or cardiopulmonary bypass times. There were no mortalities in either group. Values on table represent the Mean ± standard error.

	GIK(N=20)	NO GIK(N=20)	P Value
Postop Weight Gain (Ib)	5.8±.8	13.8±1.4	<0.0001
Cardiac Index - 12 hours postop (L/Min/M2)	2.86±0.11	2.19±.09	<0.0001
Serum Glucose - 12 hours postop (mg/ml)	135.8±8.1	238.1±9.7	<0.0001
Inotropic Score (0=no inotropes)	0.40±.15	1.25±.32	<0.0001
Time on Ventilator (hours)	8.35±.58	13.45±1.63	<0.0001
Atrial Fibrillation (%)	3(15%)	12(60%)	<0.0001
Postop Hospital Stay (days)	6.70±.34	10.20±1.48	<0.0001

CONCLUSIONS: Substrate enhancement with GIK in diabetic patients improves myocardial performance, limits weight gain, reduces inotropic dependency and the need for ventilatory support, decreases the incidence of atrial fibrillation and results in faster recovery following CABG surgery.

9:35 a.m. John Alexander Research Scholar Presentation

Richard Norris Pierson, III, Nashville, TN

9:40 a.m. Evarts A. Graham Memorial Traveling Fellowship Presentation

Christian Kreutzer, Buenos Aires, Argentina

9:45 a.m. INTERMISSION - VISIT EXHIBITS

*By Invitation

10:30 a.m. SCIENTIFIC SESSION

Ballroom, Ernest N. Morial Convention Center

Moderators: Delos M. Cosgrove, M.D.

Andrew S. Wechsler M.D.

5. Effects of Lung Volume Reduction Surgery on Pulmonary Function and Survival in Patients with End-Stage Emphysema: A Four Year Update

Michael Argenziano*, Byron Thomashow*, Patricia A. Jellen*, Lyall A. Gorenstein*, Eric A. Rose, Kenneth M. Steinglass*, Alan D. Weinberg* and Mark E. Ginsburg*, New York, New York

Discussant: David J. Sugarbaker, M.D., Boston, MA

OBJECTIVE(s): The short-term efficacy of lung volume reduction surgery (LVRS) in the palliation of end-stage emphysema has been reported. It is not clear, however, whether these effects are long-lived or whether LVRS prolongs life expectancy. In medically managed patients with FEV1 < 30% of predicted, 3-year survival has been estimated at 40%.

METHODS: Over 4 years, LVRS was performed in 144 patients with severe emphysema. Mean age was 63 ± 8 years. Preoperative PFT revealed FEV1 of 587 ± 229 cc (23 ± 8% of predicted) and dyspnea index (DI) of 3.9 ± 0.9 (range of 0-5). Operation was unilateral (n = 54) or bilateral (n = 90).

RESULTS: There were 9 perioperative deaths (6%) and 25 late deaths (2 to 34 months post-operatively). Mean followup was 29 months (range 7 to 47 months). Six months of followup was available in 133 patients, 12 months in 104 patients, 24 months in 73 patients, and 36 months in 31 patients. FEV1 and DI were improved significantly 12 and 24 months postoperatively (table), and actuarial survival at 12, 24, 36, and 42 months was 85, 74, 73, and 73%, respectively.

	FEV1 (cc)	FEV1 (% pred)	Dyspnea Index
preop	587±229	23±8	3.9±0.9
12 months	778±439*	30±14**	1.9±1.2***
24 months	784±384**	36±13*	2.0±1.9***

*p<0.05; **p<0.01;***p<0.0001

CONCLUSIONS: LVRS improves assessments of dyspnea (DI) and objective measures of pulmonary function (FEV1), and these benefits are preserved for at least two years postoperatively. Actuarial survival over 42 months after LVRS appears to be superior to reported survival for patients with end-stage emphysema managed with medical therapy alone.

*By Invitation

6. Dilation of the Pulmonary Autograft Following the Ross Procedure

Tirone E. David, Mauro P.I. De Sa*, Ahmed Omran*, Brian Sonnenberg*, Susan Armstrong*, Joan Ivanov* and Gary Webb*, Toronto, Ontario, Canada

Discussant: Nicholas T. Kouchoukos, M.D., St Louis, MO

OBJECTIVE: Dilation of the pulmonary autograft (PA) following the Ross procedure is being recognized with increasing frequency. This study was undertaken to determine the magnitude of this problem and its predictive factors.

METHODS: From 1990 to 1997, 113 patients underwent the Ross procedure. The diameters of the aortic annulus and sinotubular junction were reduced to those of the PA whenever there was discrepancy in size. There was one operative death, one early failure due to acute dilation of the PA, and one late non-cardiac death. The remaining 110 patients have had annual Doppler echocardiography to assess valve function and to measure the internal diameter of the PA sinuses. The mean follow-up was 44 ± 11 months. A stepwise logistic regression analysis was used to identify predictors of dilation of the PA.

RESULTS: No patient had more than mild aortic insufficiency (AI) early post-operatively. At the latest echocardiographic study, 62% had none or trivial AI, 33% had mild AI, and 5% had moderate AI. The diameter of the PA sinuses increased from 29.7 ± 3.7 mm (21 to 38 mm) to 33.7 ± 4.3 mm (range 26 to 46 mm), p=0.001. The PA diameter became abnormal (>39 mm) in 17 patients. PA dilation was an independent predictor of postoperative AI. Bicuspid aortic valve with large aortic annulus (>27 mm) was an independent predictor of PA dilation (odds ratio 12.5). All cases of dilation occurred in patients who had aortic root replacement with a free standing PA (p=0.01).

CONCLUSIONS: Patients with bicuspid aortic valve and dilated aortic annulus are at a high risk of PA dilation after the Ross procedure. Aortic root inclusion and subcoronary implantation of the PA may prevent dilation. Late AI is caused by dilation of the PA.

1:15 a.m. PRESIDENTIAL ADDRESS

What the Cardiothoracic Surgeon of the 21st Century Ought To Be

Lawrence H. Cohn, M.D. Boston, Massachusetts

ADJOURN FOR LUNCH - VISIT EXHIBITS

*By Invitation