WEDNESDAY AFTERNOON, May 1, 1985

1:30 p.m. Scientific Session - Grand Ballroom

38. Long Term Results After Excision of Fixed Subaortic Stenosis

H. ASHRAF*, JOSEPH COTRONEO*, NAVEEN DHAR*,

ROBERT GINGELL*, J. MICHEL ROLAND*,

DANIEL PIERONI* and S. SUBRAMANIAN

Buffalo, New York

Between 1968 and 1984, 49 patients (30 male, 19 female), ages between 0.5-22 years (mean = 7.7 ± 4.8) underwent surgical excision of fixed subaortic obstruction (discrete and tunnel) with one operative death and two late deaths. Three patients (pts.) were under one year. Forty-six (92%) were followed for 0.6 to 15 years (m = 5.6 ± 4). Twenty-five have completed 5 years and 10 are followed for 10 years or more. Twenty-five (50%) pts. had associated defects, VSD (9) being the most common. Sixteen required combined procedures. In 9 pts. subaortic obstruction was diagnosed 3 months to 8 years after a previous operation. In two other pts., the occult obstruction was discovered at surgery for VSD. There were 46 discrete sub-aortic ring (complete 31, incomplete 15) and 3 tunnel. For the discrete ring, excision alone was done in 31 while myotomy was added in 15 pts. All tunnel were treated with myotomy and myectomy. Reoperation was performed in 10 pts. (one death), 7 (15%) for recurrence and 3 for aortic valve replacement, all of whom had pre-existing aortic regurgitation which had progressed. Of the seven recurrences, 1 patient had primary surgery elsewhere. The interval of recurrence in the remainder was 2-6 (m = 4.1 ± 1.5) years, 3 recurrences in the excision alone group (31 pts.) and three in excision and myotomy group (15 pts.). Before 1976 (pre-cardioplegia) out of 22 pts., 6 had recurrence while none recurred in 26 pts. after 1976 (P = 0.0027). At 5 year follow-up, persistence of high gradient between LV and aorta after primary surgery proved to be significant. Only 3 out of 18 pts. with gradient less than 15mm recurred while all 3 pts. with gradient of 30 or higher recurred in 5 years (P = 0.015). There was no recurrence in tunnel. Thirty-nine pts. are now in NYHA Class I and 6 in Class II.

Conclusions: (1) Low early and late mortality. (2) Complete excision of the ring using cardioplegia has significantly reduced the recurrence, indicating that the recurrence may be an incomplete excision. (3) Elimination of LV-aorta gradient at initial surgery reduces recurrence rate. (4) Addition of myotomy in this series has not altered the outcome, therefore routine myotomy for discrete ring may not be necessary.

*By Invitation

39. 14 Years Experience with the Bjork Shiley Tilting Disc Prosthesis

BABULAL SETHIA*, MURDO A. TURNER*,

ROSE A. RODGER* and WILLIAM H. BAIN*

Glasgow, Scotland

Sponsored by: NICHOLAS T. KOUCHOUKOS

Birmingham, Alabama

1562 B/S tilting disc prostheses have been implanted in 1235 patients and followed for a total of 3,748 patient years.

999 standard disc (SD: 1970-1980) and 563 concavo-convex (CC: 1980-1983) prostheses were inserted. (37.3%) of patients had had previous cardiac surgery. (71.1%) were in NYHA grades III or IV.

Hospital mortality was 119/1235 (9.6%): 84/761 (11.0%) for patients with SD prostheses and 35/474 (7.3%) for CC prostheses. Late mortality was 108/1235:85/761 (11.2%) for SD prostheses, and 23/474 (4.9%) for CC prostheses.

1116 patients left hospital and have been followed for up to 14 years.

49 patients were re-operated: 26 for repair of peri-prosthetic leak: 12 for thrombotic obstruction: 4 for prosthetic endocarditis: 3 for strut fracture: and 4 for replacement of another natural valve. Mortality for re-operation was 7/49 (14.2%).

Thrombotic obstruction occurred in 20/677 SD patients (0.6/100 patient years). No case of thrombotic obstruction occurred in CC disc patients.

Systemic emboli occurred in 34/677 SD patients (1.1/100 patient years), and in 9/439 CC patients (1.3/100 patient years).

Strut fracture occurred in 3/439 CC disc patients (0.4/100 patient years): none in the SD group.

The incidence of prosthetic endocarditis (19/1116 or 0.5/100 patient years) was similar in SD and CC patients.

All patients were anticoagulated with Coumadin, complications occurred in 42 (1.1/100 patient years).

Actuarial survival at 5 years is 77.4% and at 10 years 71.9%.

Event-free survival at 5 years is 69.2% and at 10 years is 54.7%.

Our experience confirms the long term reliability of the SD prosthesis.

The incidence of thrombo-embolism with the CC disc is significantly lower, but has been offset by the 3 cases of strut fracture in this series.

*By Invitation

40. Early and Late Risk of Mitral Valve Replacement: A 12 Year Concomitant Comparison of Porcine Bioprosthetic and Disc Prosthetic Mitral Valves

LAWRENCE H. COHN, ELIZABETH N. ALLRED*,

LESLIE A. COHN*, JOHN AUSTIN*, JOSEPH SABIK*,

VERDI J. DlSESA *, RICHARD J. SHEMIN*

and JOHN J. COLLINS, JR.

Boston, Massachusetts

711 Consecutive patients (245 males/466 females, 17 to 86, 58 years) operated upon from 1/72 to 1/84 received 532 porcine bioprosthetic mitral valves (BPV) and 179 prosthetic disc mitral valves (PDV). Age, sex, and functional impairment were similar in both valve types. Mitral stenosis (MS) was the primary etiology in 363, mitral regurgitation (MR) in 348. There were 6 functional class (FC) I-II, 329 FC III and 376 FC IV patients. Associated procedures were done in 253/711 (36%) (168 [24%] received CABG). Overall operative mortality was 76/711 (10.6%), 34/476 (7%) for MVR alone, 30/168 (18%) for MVR + CABG (p<.0001), 35/363 (9.6%) for MS, 41/348 (11.7%) for MR, 50/532 (9.3%) for BPV and 26/179 (14.5%) for PDV, 13/535 (3.8%) FC I-III, 63/376 (16.7%) FC IV.

Long-term follow-up for 482 BPV and 153 PDV patients was 6-151 mos (50 mos); 52 (8%) were lost to follow-up. Overall actuarial survival at 108 mos was 60 ± 3%, for MS 67.5 ± 4%, MR 53 ± 5% (p=.0001), for BPV 67 ± 4%, 41 ± 6% for PDV (P = <.0001), for MVR67 ± 4%, for MVR + associated procedure 45 ±6% (p<.0005), for FC I-III 71 ± 4%, and 51 ± 4% for FC IV (p<.0001). Thromboembolic events (TE), including thrombosis, occurred overall in 68 patients (2.8%/pt yr); 2.43 for BPV and 4.37 for PDV (p<.05). Actuarial freedom at 108 months from TE overall was 83 ± 2%, for BPV 84.5 ± 3, for PDV 78 ± 6 (p = NS). For BPV patients in atrial fibrillation (AF), probability freedom from thromboemboli at 12 years was 80 ±4% versus 92 ± 2% for sinus rhythm (SR).

Primary valve dysfunction occurred in 18 patients with BPV, 3 with PDV. Freedom from primary valve dysfunction at 108 mos for BPV was 89 ± 3% and 93 ± 4% for PDV (p = NS), (0.9%/pt yr for BPV, 0.6 for PDV). There were 18 perivalvular leaks, but no significant difference in patients according to valve (0.7/pt yr BPV vs 0.8 for PDV). The probability of freedom from late endocarditis at 108 mos for BPV was 89 ± 3% and 95 ± 3% for PDV(p = NS).

Early and late survival after MVR reflects preoperative functional class and associated coronary artery disease. Thromboembolism and valve thrombosis are major risks of PDV, although there is some element of valve dysfunction. Primary valve dysfunction is a major risk factor of mitral BPV and compared to aortic valve data, primary valve dysfunction is higher, but the risk of endocarditis may be lower. The risk of TE in BPV patients with chronic atrial fibrillation is significant.

*By Invitation

41. Comparison of Bioprosthetic and Mechanical Valve Replacement for Active Endocarditis

GEORGE J. REUL, JR., MICHAEL S. SWEENEY*,

DENTON A. COOLEY, DAVID A. OTT,

J. MICHAEL DUNCAN*, O. HOWARD FRAZIER*

and JAMES J. LIVES AY* Houston, Texas

The choice between bioprosthetic valve (BPV) or mechanical prosthetic valve (MPV) replacement for active valvular endocarditis has been controversial. To establish the role of each, we reviewed 185 patients who underwent valve replacement for active valvular endocarditis during the past 5 years. All patients had life-threatening, active bacterial endocarditis of a native or prosthetic valve. The BPV group (Group I, 88 patients) had replacement with the lonescu-Shiley pericardial valve and the MPV group (Group II, 97 patients) with the St. Jude Medical valve. The male/female distribution, age range, and functional classification were the same in both groups. Mean follow-up was 22.3 months, and all events occurred within the first two years in both groups. Valve replacement was done because of native valve endocarditis in 76 patients in Group I and 49 patients in Group II (p>0.01). Of the remainder of Group I patients, 6 had endocarditis of a BPV and 6 of a MPV; of the remainder of Group II patients, 30 had endocarditis of a BPV and 18 of a MVP (p>0.01). Early mortality was not significantly different (14 in each group). Of the 74 survivors in Group I, 15 underwent valve reoperation: 9 because of recurrent endocarditis and 5 because of sterile perivalvular leakage. This was significantly different (p<0.01) from Group II, where only 5 patients underwent valve reoperation: 4 for recurrent endocarditis and one for sterile perivalvular leakage. The actuarial rate for freedom from reoperation was also significantly higher in Group II patients, where 93.4% were free from reoperation at 3 years compared to 75% at 4 years in Group I patients. Two thromboembolic events occurred in both groups in the first year of follow-up. Actuarial survival, which was not significantly different, was 78.7% at 4 years in Group I and 82.6% at 3 years in Group II. Patients receiving BPV for active endocarditis had a significantly higher reoperation rate and a significantly greater incidence of recurrent endocarditis; therefore, we prefer the MPV for valve replacement in most patients who have active endocarditis.

*By Invitation

42. Prospective Evaluation of Mediastinoscopy for Assessment of Carcinoma of the Lung

WILLIAM P. LUKE*,

FREDERICK GRIFFITH PEARSON,

THOMAS R.J. TODD*,

GEORGE ALEXANDER PATTERSON*

and JOEL DAVID COOPER

Toronto, Ontario, Canada

Between 1979 and 1984 cervical mediastinoscopy was carried out on 960 (61.5%) of the 1559 patients admitted to the Thoracic Surgical Service of the Toronto General Hospital with a diagnosis of lung cancer. In 127 cases, concomitant anterior mediastinostomy was also performed. Positive nodes were found in 286 patients (30%). There was an overall 2% complication rate with no mortality. Positive mediastinal nodes were found in 17% of squamous carcinomas, 23% adenocarcinomas, 49% small cell, 30% large cell undifferentiated and 9% bronchoalveolar. Positive mediastinal nodes were found with equal frequency in right and left sided tumours and occurred in 34% of tumours of the main bronchus, 26% of upper lobe tumours and 17% of lower lobe tumours. Of the 674 negative mediastino-scopies, 562 patients (83%) came to thoracotomy, at which time 58 patients (10%) were found to have positive mediastinal nodes. Overall resectability rate was 93% (86% curative, 7% palliative). Twenty per cent of the resections were pneumonectomies. Fifty-five (19%) of the 268 positive mediastinoscopy patients were selected for thoracotomy. In this group, resectability rate was 85% (67% curative, 18% palliative). Pneumon-ectomy rate was 34%. Follow-up data is complete for 925 of the 960 patients with a mean follow-up of 36 months.

We conclude that routine mediastinoscopy can be done with negligible morbidity and provides essential information for the management of patients with lung cancer.

*By Invitation

43. Comparative Merits of Conventional, CT and MR Imaging in Assessing Mediastinal Involvement in Surgically Confirmed Lung Carcinoma

NAEL MARTINI, ROBERT HEEL AN*,

JACK WESTCOTT*, MANJITBAINS,

PATRICIA MCCORMACK*, JAMES CARAVELLI*,

ROBIN WATSON* and MUHAMMAD ZAMAN

New York, New York

Nineteen patients presenting with potentially resectable malignant tumors of the lung had computed tomography (CT) and magnetic resonance (MR) imaging of the chest in addition to PA and lateral chest x-rays and bronchoscopy. The purpose of the study was to assess the extent of tumor involvement in hilum and mediastinum by direct invasion and by regional lymph node metastasis.

At thoracotomy a mediastinal lymph node dissection or sampling was carried out to correlate nodal involvement with the preoperative studies. All nodes were examined histologically and their size and location recorded.

The tumor was peripheral in 13 patients and central in 6. Histologically, 11 were adenocarcinomas, 6 squamous cancers and 2 atypical carcinoids. Preoperatively, 6 were recorded to have NO disease, 6 N1, and 7 N2. Pathologically, 6 were NO, 3 N1 and 10 N2.

As expected, conventional imaging correlated poorly with the findings at surgery with false positives in 2 and false negatives in 9 patients.

CT and MR imaging demonstrated N2 disease in 9 and 10 N2 patients. False positive N2 disease was reported by both methods in 4 patients, 2 of whom had enlarged hyperplastic lymph nodes.

MR imaging was accurate in assessing the hilum in all 9 patients with NO and N1 disease whereas 2 false positives and 1 false negative N1 were reported on CT.

Hilar and mediastinal disease was shown with greater clarity in MR imaging but no advantage of MR over CT could be demonstrated in detecting N2 disease or mediastinal involvement. Neither method could differentiate hyperplastic from mediastinal nodes.

3:30 p.m. Adjournment

*By Invitation