WEDNESDAY MORNING, APRIL 29, 1992

7:30 a.m. FORUM SESSION II - General Thoracic Surgery

Los Angeles Ballroom

F11. Improved Ultrastructural Lung Preservation With Prostaglandin E1 As Donor Pre-Treatment in a Primate Heart Lung Transplant Model

ROBERT S. D. HIGGINS, M.D.*, GEORGE V. LETSOU, M.D.*,

JUAN SANCHEZ, M.D.*, RICHARD EISEN, M.D.*,

KENNETH L. FRANCO, M.D.*,

GRAEME L. HAMMOND, M.D. and

JOHN C. BALDWIN, M.D.

New Haven, Connecticut

Donor pre-treatment utilizing Prostaglandin E₁ (PGE1) as a pulmonary vasodilator has developed as a simple, effective means to provide excellent preservation in heart-lung transplantation. This study was undertaken to investigate the Ultrastructural preservation of the lung using prostaglandin El and other pulmonary vasodilators in a primate heart-lung transplantation model.

Methods: Heart-lung transplantation was performed in 14 African green monkeys. Donor cardiac preservation was achieved with cold crystalloid car-dioplegic solution (10 ml/kg). Lung preservation was achieved with cold, modified Euro-Collins solution (8meq MgSO4 + 65 cc 50% Dextrose added to standard solution) delivered into the main pulmonary artery (60ml/kg/ total). Vasodilator agents were administered systemically 15 minutes prior to aortic cross-clamping. Central venous, femoral and pulmonary arterial pressures were monitored during infusion. The heart-lung grafts were stored at 4°C for 6 hrs. Three groups of animals were studied: 5 donors with PGE1 (0.1 to 4.0 meg/kg/min), 5 donors with prostacyclin (0.1 to 0.35 meg/kg/min), and 4 donors with nitroprusside (0.8 to 5.0 mcg/kg/min). After transplantation, arterial blood gases and lung biopsies were obtained at 1-3 hrs. and at conclusion of the study. Five formalin blocks per specimen were sectioned for hematoxylin and cosin staining. Cellular architecture and endothelial cell swelling were evaluated using electron microscopy. The specimens were graded for alveolar hemorrhage, endothelial cell swelling (grade 1 = minimal to grade 3 = severe) and preservation of cellular architecture and a mean score was obtained for each preservation agent.

Results: Endothelial cell swelling correlated with the degree of cellular preservation. PGEi specimens demonstrated the least amount of endothelial swelling (mean score .8) compared to PCI and NTP specimens (mean score 1.4 and 2.7 respectively). All NTP specimens demonstrated moderate to severe endothelial cell swelling. Interstitial and alveolar hemorrhage was noted in poorly preserved specimens, but there were no significant differences between groups.

Conclusions: Prostaglandin El provides improved cellular preservation by decreasing the extent of endothelial cell swelling on electron microscopy.

*By Invitation

F12. The Effect of Corticosteroids on Bronchial Blood Flow After Lung Transplantation

KENJI INUI, M.D.*, H. J. SCHAFERS, M.D.*,

VERA BECKER, M.D.* BIRTE ONGSIEK, M.D.*,

AXEL HAVERICH, M.D.* and HANS G. BORST, M.D.

Hannover, Germany

The prophylactic administration of corticosteroids has been discussed controversially in clinical lung transplantation for fear of an increased prevalence of bronchial complications. In an experimental investigation, the effect of prednisolone on bronchial blood flow was investigated, using a model of modified unilateral lung transplantation in pigs. The bronchial anastomosis was created between the donor trachea and recipient left main bronchus. Im-munosuppression consisted of cyclosporine (15 mg/kg/d) and azathioprine (2 mg/kg/d) in group I (n = 6); in group II (n = 6), prednisolone (1 mg/kg/d) was added. The animals were studied on the 7th postoperative day. Bronchial blood flow was estimated using laser doppler velocimetry (LDV) and radioisotopic studies (RI). Segments of the graft airway and parenchyma were assessed for signs of ischemia or rejection.

In group I, bronchial blood flow (in percent of reference flow) at the graft main carina was markedly reduced (LDV: 34.9 ± 5.8%; RI 33.3 ± 19.8%). By comparison, blood flow in group II was significantly improved (LDV: 55.3 ± 3.9%, p<0.005; RI: 57.1 ± 18.3%, p<0.05). Histologic classification of acute pulmonary rejection resulted in similar scores in both groups. In group I, necrotic changes at the graft main carina were observed in 4 of 6 cases with extension into the muscular and cartilagenous portions of the airway wall. By comparison, only limited loss of epithelium with preservation of the deeper layers of the airway were seen in all cases of group II.

Addition of prednisolone to immunosuppression with azathioprine and cyclosporine results in decreased ischemia and significantly improves bronchial blood flow after lung transplantation. Prophylactic administration of corticosteroids is expected to alleviate bronchial ischemia in clinical lung transplantation and thus reduce the incidence of postoperative airway complications. This is confirmed by our clinical experience with 46 lui.£ transplantation.

*By Invitation

F13. Re-establishment of Lymphatic Drainage After Lung Transplantation in Canine Model

RENATO RUGGIERO, M.D.*, JAROSLAW MUZ, M.D.*,

ROBERTFIETSAM, M.D.*,

GREGORY A. THOMAS, M.D.*, ROBERT J. WELSH, M.D.*,

JAMES E. MILLER, M.D.*,

LARRY W. STEPHENSON, M.D. and

FRANK A. BACIEWICZ, JR., M.D*

Detroit, Michigan

The lymphatic channels between the lung and the mediastinum are interrupted during lung transplantation. Although clinical and radiologic impressions suggest that lymphatic connections are re-established in the postoperative period, this has not been proven. The purpose of this investigation is to document lymphatic regeneration after lung transplant. A first group of 6 control dogs had transthoracic injection in the periphery of the left upper and lower lobes of 0.2 ml of antimony sulphide colloid (ASC) tagged with Technetium-99m (Tc-99m). ASC is a particle 2-15 nm in size which is absorbed only through lymphatic channels and concentrated in the tributary lymph nodes. 24 hours after injection the animals underwent scintigraphic studies and two-dimensional images of the lung and mediastinum. A second group of 4 dogs underwent division and reanastomosis of the left main bronchus. The left pulmonary artery and veins were dissected from all surrounding tissue. On the second postoperative day and then weekly for 4 weeks, the animals underwent Tc-99m-ASC injection and were studied 24 hours later as above. A third group of 4 dogs underwent left lung allotransplant using standard techniques including immunosuppression with CSA 50 mg/kg/day and AZA 1 mg/kg/day. The animals were studied with Tc-99m-ASC injection and lung lymphoscintigraphy every week for 6 weeks. All control dogs demonstrated mediastinal lymph nodes at 24 hours. 3 of 4 dogs in the reimplantation group showed mediastinal lymph nodes in all studies after 8 days. In the transplant group, mediastinal nodes were visualized for the first time 2 to 4 weeks after the surgery, and at every study from then on. In conclusion, lung lymphoscintigraphy with transthoracic injection of Tc-99m-ASC is a reliable technique for the study of pulmonary lymphatic flow. This experiment conclusively shows for the first time that lymphatic drainage after lung transplantation is re-established beginning as early as the second postoperative week.

*By Invitation

F14. Growth Potential of Reduced Size Mature Pulmonary Lobar Transplants

JOHN A. KERN, M.D.*, IRVING L. KRON, M.D.,

CURTIS G. TRIBBLE, M.D.*, TERRY L. FLANAGAN, MPH*,

BARRYB.K. CHAN, M.D.* and

WALTER W. SCOTT, B.S.*

Charlottesville, Virginia

The use of mature pulmonary lobes for pediatric lung transplantation has recently been described and represents a potential avenue for solving the severe pediatric donor lung shortage. To provide adequate long term function, however, transplanted mature lobes need to grow in proportion to the recipient. The total number of alveoli in the human lung reaches adult levels by 8-10 years and in the pig by 12 weeks. Whether or not an already mature lobe can grow by forming new alveolar units, following transplantation into a developing recipient, is not known. To study the growth potential of mature lobar transplants, we measured functional residual capacity (FRC), fixed lung volume, alveolar airspace percent, alveolar size, and total number of alveoli in isolated mature left lower lobe lung transplants 12 weeks following implantation into growing piglets. Comparisons were made to age matched control left lower lobes (studied immediately after a left upper lobectomy) to determine if functional or morphologic growth had occurred. The transplanted lobes, and lobes which served as controls were all explanted from 6 month old animals. Helium dilution was used to determine FRC and standard morphologic techniques utilizing the point counting method were used for determination of alveolar airspace percent, and alveolar numbers. Recipients of the transplanted lobes were 9 weeks old and weighed 20 ± 1 kilograms initially. By the end of the 12 week holding period, the animals had increased their body weight by approximately 4 times (88 ±2 kg). Results: values are means ± SEM:

Lobe	FRC (cc)	Fixed Lung Volume (cc)	Alveolar Airspace Percent	Alveolar Size (Diameter in um)	Alveolar Number (x106)
Transplanted (n = 5)	579 ± 42	914 ± 50	72 ± 3	72 ± 3	112 ± 11
Control (n = 6)	570 ± 89	685 ± 35*	74 ± 2	71 ± 2	96 ± 9

*p<0.05 by unpaired t-test

No statistically significant differences were seen in FRC or morphologic analysis of alveoli in the transplanted versus control lobes. Conclusions: Transplantation of mature lobes into growing animals did result in significant growth of the lobes as determined by fixed volume, though not an increase in alveolar number of functional capacity. Reduced size mature lobar transplantation results in compensatory lobar growth which is not due to a statistically significant increase in functional gas exchanging units.

*By Invitation

F15. Allograft Replacement of the Trachea: Experimental Synchronous Revascularisation of Composite Thyro-Tracheal Transplantation

JOSEPH F. KHALIL-MARZOUK, M.D.*,

PETER COLOSTRAW, FRCS* and

MERLY GRIFFITHS, MRCPath*

Taif, Saudi Arabia

Sponsored by: Joel D. Cooper, M.D., St. Louis, Missouri

The purpose of this study is to assess the structural integrity of the transplanted trachea following composite Thyro-Tracheal allograft replacement in dogs, utilizing microvascular anastomoses based on the blood supply of the related thyroid gland.

Twelve-ring segments of the cervical trachea and the thyroid gland were resected in 18 Beagle dogs. They were grouped as follows: 6 dogs underwent non-vascularised transplants (Group A), 12 dogs received revascularised Thyro-Tracheal composite allografts establishing the anastomoses of the cranial thyroid arteries to the ipsilateral common carotid arteries, 6 dogs did not receive any immunosuppressive agents (Group B) and 6 dogs were given Cyclosporin A in the dose of 25 mg/kg body weight (Group C). Transplant operations were performed in pairs of dogs in the form of double swap procedures, half were implanted in a fresh state in all 3 groups and half were preserved in cold Hartman's solution at 4°C for a period averaging 3 hours.

ALL animals survived the operation, they were sacrificed at intervals (at Humane end-point) between 3 and 28 days. Gross anatomy and light microscopic studies were performed in all cases. Group A demonstrated total loss of structural integrity of the tracheal cartilages and soft tissues as early as 3 days in all animals. Group B showed adequate preservation of viability of tracheal cartilages but necrosis presumed rejection of the soft tissues. Group C showed preservation of tracheal cartilages and soft tissues in all but one case, there was no evidence of cartilage necrosis in the form of empty lacunae, the tracheal muscles and the thyroid gland remained histologically intact, however the lining mucous membrane was colonised by microorganisms and was partially disrupted but there was evidence of early epithelial regeneration at the tracheal anastomoses.

We concluded that revascularisation of the transplanted trachea using the thyroid arteries does maintain the vascularity and hence viability of the tracheal cartilages enabling a reliable substitute for long segment tracheal resections. We predict that the clinical success of this technique will solve a major problem in surgery of the airways.

*By Invitation

F16. Lazaroid U74500A as an Additive to UW Solution for Pulmonary Grafts in the Rat Transplant Model

RYO AEBA, M.D.*, WILLIAM A. KILLINGER, M.D.*,

ROBERT J. KEENAN, M.D,*, SAMUEL A. YOUSEM, M.D.*,

ROBERT L. HARDESTY, M.D. and

HARTLEY P. GRIFFITH, M.D.

Pittsburgh, Pennsylvania

Lazaroids, a class of novel 21 amino-steroids, have been reported to be potent inhibitors of iron-dependent lipid peroxidation, which is a major contributing factor to ischemia-reperfusion injury in lung. The aim of this study was to determine the preservation effect of a lazaroid U74500A for pulmonary grafts.

The pulmonary arteries of Lewis rats were flushed with either cold (0 °C) standard University of Wisconsin solution (UW) or UW with 30µM of U74500A substituted for the dexamethasone (UW-L) and then the heart-lungs blocks were stored in the same solution. After 6 or 12 hours of cold storage at 0 °C the left lungs were orthotopically transplanted into isogeneic recipient rats and reperfused for 1 hour. Pulmonary function was assessed by measuring oxygen (PaO₂) and carbon dioxide (PaCO₂) tensions in arterial blood on 100% oxygen after removal of the right lung. Tissue lipid peroxide (LP) levels in the transplanted lungs were measured as a thiobarbiturate-acid reactive substances. There were 6-8 animals in each group. The results were expressed as mean ± SE.

	UW (6 hours)	UW-L (6 hours)	UW (12 hours)	UW-L (12 hours)
PaO2 (mmHg)	309 ± 81	436 ± 30	27 ± 3	339 ± 70*
PaCO2 (mmHg)	28.2 ± 2.3	24.4 ± 3.8	47. 7 ± 7.0	24.3 ± 2.7**
LP (µmol/g)	.88 ± .07	.54 ± .07*	1.30 ± .09	.69 ± .07**

*:p<0.01 **:p<0.001 vs UW

We conclude that U74500A in the flush and storage solution enhances the preservation of the pulmonary graft in this model.

*By Invitation

F17. NR-LU-10 Monoclonal Antibody Scanning: A Helpful New Adjunct to CT in Evaluating Non-small Cell Lung Cancer

HOMER MACAPINLAC, M.D.*, VALERIE W. RUSCH, M.D.*,

ROBERT HEEL AN, M.D.*, ELISSA KRAMER, M.D.*,

STEVEN LARSON, M.D.* and

ROBERT J. GINSBERG, M.D.

New York, New York

CT scanning has improved noninvasive staging of lung cancer patients, but has the deficiency of not distinguishing benign from malignant lesions. This prospective trial evaluated the usefulness of a new monoclonal antibody (MoAb) imaging technique as an adjunct to CT by assessing (1) its clinical applicability; (2) its accuracy in detecting malignancy in primary lung tumors and mediastinal nodes. The MoAb, NR-LU-10 (NeoRx corporation), is a murine IgG2b Ab labeled with Tc-99m which recognizes a 40 kD glycopro-tein expressed in non-small cell lung cancers (NSCLC). Methods: (1) patients (pts) with potentially resectable NSCLC were eligible; (2) all pts had contrast-enchanced chest CT scans; (3) 25-30 mCi of MoAb were infused I.V. followed by whole body and SPECT imaging 14-17 hours later; (4) subsequent mediatinoscopy or thoracotomy with complete mediastinal nodal mapping provided pathologic correlation.

Results: 24 pts entered. Men:Women = 14:10. No allergic reactions or other side effects of MoAb were seen. Interference from prior V/Q scan precluded adequate imaging in 1 pt, but high quality MoAb images were obtained in the other 23 pts (96%). There was no background activity in lung, and little in the mediastinum. Pathologic correlation:

	MoAb Uptakein 1° tumor	Mediastinal MoAb	Nodal Involvement CT
True(+)	22	8	9
True(-)	1*	7	7
False (+)	0	6	6
False (-)	0	1	1

* = 1 ° lung lesion subsequently proven to be inflammatory

Conclusions: (1) NR-LU-10 MoAb scanning is safe, is easily performed, and produces high quality images of lung and mediastinum; (2) it is very accurate in detecting 1 ° NSCLC; (3) it is as accurate as CT in assessing involvement of mediastinal nodes. NR-LU-10 MoAb holds promise as an adjunct to CT by distinguishing benign from malignant lung lesions.

*By Invitation

F18. The Prognostic Significance of Flow Cytometry in Lung Cancer

THOMAS W. RICE, M.D.*, THOMAS W. BAUER, M.D.*,

GORDON N. GEPHARDT, M.D.*,

SHARON V. MEDENDORP, MPH*,

DENISE A. McLAIN, B.S.* and

THOMAS J. KIRBY, M.D.*

Cleveland, Ohio

Sponsored by: Delos M. Cosgrove, M.D., Cleveland, Ohio

The prognostic significance of stage in lung cancer is well known but the significance of other factors, including DNA ploidy analysis, is unclear. To clarify the value of DNA ploidy analysis, we prospectively studied single parameter flow cytometry of fresh tissue from 278 patients with resected primary lung cancer from whom adequate tissue was present. These results as well as age, sex, cell type, histologic grade, and AJCC stage were correlated with survival.

The mean age of the patients was 65.4 years; 65% were male. Cell types were: adenocarcinoma 107 (38.6%); squamous cell 100 (36%); large cell 56 (20%); adenosquamous 8 (3%); small cell 6 (2%); and giant cell (0.4%). Histologic grades were: I (well differentiated) 15 (5%); II100 (36%); and III 163 (59%). AJCC stages were: I 154 (56%); II 39 (14%); III 76 (27%); and IV 9 (3%). Mean follow-up was 18.9 months (range <1-59). One hundred eighty-one patients (65%) were alive at last follow-up. Survival at one year was 75% (±3%) and at three years survival was 53% (±6%).

Each flow cytometry histogram was classified as follows: 1) no detectable aneuploidy 48 (17%); 2) hyperdiploid 152 (55%); 3) hypodiploid 4 (1%); 4) hypertetraploid 16 (6%); 5) multiple aneuploid populations 49 (18%); and 6) multiple aneuploid populations with one hypertetraploid population 9 (3%). For tumors with no detectable aneuploidy, mean S-phase was 4% (range 1-12%) and mean S + G2M phase was 9% (range 0.3-34%).

Multivariate analyses using the Cox Proportional Hazards Model for survival showed that 1) increasing AJCC stage (p<0.001) and male sex (p = 0.007) were the only factors of independent (negative) prognostic significance and 2) neither the presence nor absence of DNA aneuploidy (p = 0.71), the classification of DNA histogram (p = 0.81), nor the results of cell cycle analysis (S phase:p = 0.83 and S + G2M:p =0.47) showed significant correlation with survival in this group of patients.

*By Invitation

F19. Assessment of Impaired Diaphragmatic Function After Thoracotomy

JOHN C. WAIN, M.D.*,

MARIE-DOMINQUE FRATACCI, M.D.*,

WILLIAM K. KIMBALL, M.D.* and M. ZAPOL, M.D.*

Boston, Massachusetts

Sponsored by: Hermes C. Grillo, M.D.,

Boston, Massachusetts

Experimental and clinical studies have suggested that diaphragmatic function is impaired after pulmonary resection. Previous clinical studies have assessed diaphragmatic function indirectly. However, direct, reliable diaphragmatic length measurements can be obtained with implantable sonomicrometry crystals. To directly assess diaphragmatic function in humans, sonomicrometry crystals were implanted in the costal diaphragm after thoracotomy and pulmonary resection in 6 patients (4 right thoracotomy, 2 left thoracotomy). Following placement of an esophageal balloon catheter and recovery from anesthesia, measurements of resting diaphragmatic muscle length and percent shortening, changes in trans-diaphragmatic pressure (Pdi) and in tidal volume (Vt) were made during mechanical ventilation and during spontaneous respiration before and after a dose of epidural anesthetic (BEp, PEp, respectively).

	mechanical	spontaneous
		BEp	PEp
resting length	16.3 ± 2.2	15.0 ± 1.9	15.5 ± 1.8
% shortening	7.9 + 3.0	1.6 ± 1.6*	-.9 ± 2.6
Pdi	-5.4 ± 1.6	7.7 ± 1.5**	11.5 ± 1.9#
Vt	458 ± 102	390 ± 78	555 ± IStttt

*p < 0.05 versus mechanical, **p < 0.01 versus mechanical

#p< 0.05 versus BEp, ##p < 0.01 versus BEp

Conclusions: Diaphragmatic shortening is significantly impaired during spontaneous respiration post-thoracotomy. Although epidural anesthesia significantly improves spontaneous tidal volume, this is not due to increased diaphragmatic shortening.

*By Invitation

F20. Barrett's Esophagus - A Functional Foregut Disorder?

TOM R. DeMEESTER, M.D. and

HUBERT J. STEIN, M.D.*

Los Angeles, California

Understanding the pathophysiology of Barrett's esophagus will provide better use of subject therapy to treat this complicated reflux problem. To understand the development of Barrett's esophagus in patients with gastroesophageal reflux disease we compared symptoms (severity score 0-3), the gastric secretory state, esophageal acid exposure (pH < 4, < 3, and < 2), lower esophageal sphincter (LES) function, and circadian esophageal motor activity in 15 patients with Barrett's esophagus to 24 patients with esophagitis, and 25 normal volunteers.

Results: Compared to patients with esophagitis, patients with Barrett's esophagus had less heartburn and regurgitation, but an increased frequency and duration of reflux episodes (pH < 4), and % time pH < 4, pH < 3, and pH < 2 in the esophagus (table). This was associated with an increased basal and maximum gastric acid output (BAO and MAO), decreased LES pressures, and increased frequency of ineffective contractions in the distal esophagus (i.e., isolated contractions or contractions < 30 mmHg) on 24-hour ambulatory esophageal motility monitoring (table).

Table:	Normals	Esophagitis	Barrett's
Heartburn	0 ± 0	2.3 ± 0.2	1.4 ± 0.3*
Regurgitation	0 ±	2.2 ± 0.3	1.6 ± 0.2*
# reflux episodes	21.2 ± 1.9	72.5 ± 6.9	110.2 ± 12.4
# reflux episodes >5 minutes	1.3 ± 0.2	8.1 ± 1.2	15.9 ± 2.2*
%time pH < 4	2.5 ± 0.6	10.3 ± 2.9	24.5 ± 4.8*
% time pH < 3	1.2 ± 0.2	3.4 ± 0.4	10.2 ± 1.9*
%time pH < 2	0.4 ± 0.2	2.1 ± 0.5	4.9 ± 0.9*
BAO	-	2.7 ± 0.4	6.3 ± 1.3*
MAO	-	13 ± 3	23 ± 5*
LES pressure	12.1 ± 2.1	8.2 ± 1.7	4.3 ± 0.7*
% ineffective contr.	18.7 ± 2.3	22.9 ± 4.2	47.9 ± 7.8*

Means ± SEM, *:p < 0.01 vs patients with esophagitis

Conclusion: Despite less symptoms patients with Barrett's esophagus have a markedly increased esophageal exposure to higher concentrated gastric juice compared to patients with esophagitis. This appears to be due to an increased frequency and prolonged duration of reflux episodes with concentrated acid secondary to a combination of acid hypersecretion with amechanically defective lower esophageal sphincter and inefficient esophageal clearance function. Planned surgical therapy should reduce esophageal acid exposure to normal by improving sphincter function with a Collis gastroplas-ty and a partial fundoplication so as not to potentiate the detrimental effect of reduced body function.

*By Invitation

WEDNESDAY MORNING, April 29, 1992

9:00 a.m. SIMULTANEOUS SCIENTIFIC SESSION D - ADULT CARDIAC SURGERY - Los Angeles Ballroom

39. The Right Gastroepiploic Artery Graft: Clinical and Angiographic Mid-Term Results in 200 Patients

HISA YOSHI SUMA, M.D.*, YASUHIKO WANIBUCHI, M.D.*,

YASUSHI TERADA, M.D.*, SACHITO FUKUDA, M.D.*,

TETSURO TAKA YAMA, M.D.* and

SHOICHIFURUTA, M.D.

Tokyo, Japan

Sponsored by: Donald B. Doty, Salt Lake City, Utah

From March 1986 to September 1991, the right gastroepipicic artery(GEA) graft has been used for CABG in 200 patients (171 males, 29 females, mean age 58 ranged from 6 to 80) and they were followed up from 2 to 66 months with a mean of 23 months.

There were 16 reoperation cases and 176 patients had triple vessel or left main disease. GEA (182 in-situ and 18 free grafts) was anastomosed to 11 anterior descending, 3 diagonal, 26 circumflex and 160 right coronary arteries. The internal thoracic artery was concomitantly used in 192 patients and mean number of distal anastomoses was 3.3 including additional saphenous vein grafts. Postoperative angiography was performed in 143 patients within 6 postoperative months (mean 2 months) and second angiography was made at 1 to 3 postoperative years (mean 2 years) in 38 patients sequentially.

There were 6 early and 3 late deaths. New Q wave was noted in 4 patients. Duration of surgery and postoperative complication did not increase with use of GEA. Relief of angina was noted in 196 patients. GEA patency was 95% (136/143) at early and 95% (36/38) at late potoperative period. There were 4 successful PTCA through in-situ GEA graft for GEA anastomotic stenosis. In stress myocardial scintigraphy sequentially performed at preoperative, and immediate, 1 year and 2 year postoperatively in 11 patients, washout rate (%)of GEA grafted area improved from 35 ± 10 to 45 ± 15 (P<0.05) and was maintained in 43 ± 6 and 48 ± 9 at respective periods. By implantable dop-pler flow study, in situ GEA graft flow increased with meal.

In conclusion, GEA is a suitable arterial conduit for CABG in terms of low surgical risk, high patency rate and excellent patient's outcome.

*By Invitation

40. Aorto-Coronary Bypass Graft Patency After High Dose Aprotinin

BENJAMIN P. BIDSTRUP, FRACS*

London, England

Sponsored by: J.C.R. Lincoln, FRCS, London, England

High dose aprotinin has been shown in several single centre studies to dramatically reduce postoperative blood loss. Remarkedly few side effects of this drug have been reported in these studies and in reports of routine clinical use. However concern has been expressed that the use of haemostatic agents during or after aorto-coronary bypass graft (ACBG) surgery might affect patency of vein grafts. The aim of this prospective study was to determine if high dose aprotinin resulted in increased vein graft occlusion after primary ACBG.

Patients and Methods: In a placebo controlled, double blind study, 96 male patients (mean age 58.9 years) received a total of 363 bypass grafts. High dose aprotinin (6 x 10^{^}6 kallikrein inhibitory units approximately) was administered intra-operatively to 47, the remainder receiving placebo. The operative procedure, bypass and postoperative care was maintained uniformly within limits imposed by the clinical course. All procedures were carried by a single surgical service. Vein graft patency was assessed at 6 - 12 days postoperatively by magnetic resonance imaging, using a spin-echo sequence in 90 (93%) of the patients. Both groups were similar with respect to age, weight, bypass time and number of grafts placed. All patients received low dose aspirin beginning the first postoperative day. Patency of vein grafts to endarterectomized arteries was analyzed separately.

Results: At the conclusion of the study, 269 vein grafts were assessed, 131 in the aprotinin group and 138 in the placebo group. In the aprotinin group, 96.2% of grafts were patent vs 97.1% in the placebo group (70% CL 93.7 -97.8% vs 94.8 - 98.5%). This difference was not significant when tested using the ratio estimate procedure, assuming non-independence of grafts (p > 0.05). All grafts were patent in the aprotinin group in 86.1 % of patients compared with 89.4% (70% CL: 78.6 - 91.5% vs 82.8 - 93.9%) in the placebo group. Postoperative hemoglobin loss was reduced by 50% in the aprotinin treated patients and homologous blood transfusions by 45%.

Conclusions: Early graft occlusion is due to thrombosis and is determined by the quality of the artery and its run-off, the quality and type of the conduit used, the surgeon and the technical adequacy of the anastomoses, and the interaction of the coagulation system. Early (7-10 day) vein graft patency, as determined by MR imaging is not adversely affected by high dose aprotinin use. At the same time, similar reductions in blood loss and blooduse as have been previously described, were achieved. The improvement in haemostasis afforded by aprotinin is not achieved by the creation of a pro-thrombotic state. Rather, haemostasis after cardiopulmonary bypass is less disturbed, thereby reducing the bleeding tendency.

*By Invitation

41. Left Ventricular Function in Experimental Mitral Regurgitation With Intact Chordae Tendineae

HANI A. HENNEIN, M.D.*, MICHAEL JONES, M.D.,

CHRISTOPHER D. STONE, M.D.* and

RICHARD E. CLARK, M.D.

Bethesda, Maryland

Left ventricular (LV) function in mitral regurgitation (MR) has typically been studied in models that either sever the chordae tendineae or create a ventriculo-atrial shunt. These methods may have adverse effects on LV function independent of the regurgitant lesion itself. An animal model of chronic MR was therefore developed that both preserves annulo-ventricular continuity and avoids the use of external shunts. A circular 0.16-0.24 mm/kg defect was created in the anterior mitral valve leaflet of weaning sheep under direct vision using cardiopulmonary bypass. Six animals were studied pre- and immediately postoperatively (AMR-C and AMR groups, respectively), and 20 animals were studied 8.1 ±0.2 (mean±SD) months postoperatively (CMR group). CMR animals were compared to an age and weight matched, normal, nonoperated control group (CMR-C, n = 7). Volumetric data were derived from digitalized cineangiographic images, and Emax calculated from pressure-volume loops obtained from the simultaneous recording of LV pressure by micromanometer tipped LV catheters, and volumes obtained from digitalized images of epicardial echocardiographic recordings. The data showed that there was severe MR in both the AMR and CMR groups, with a calculated regurgitant fraction of 37 ±7%.

GROUP	CI	LV dP/dt	EDV	ESV	EF	Emax
AMR-C	102 ± 19	1567 ± 432	110 ± 17	45 ± 7	59 ± 6	11.8 ± 1.2
AMR	113 ± 26*	1376 ± 245	121 ± 23*	46 ± 10	62 ± 7	11.4 ± 1.7
CMR-C	114 ± 29	1588 ± 334	1456 ± 34	63 ± 20	57 ± 9	10.3 ± 0.9
CMR	72 ± 28b	806 ± 293b	202 ± 32b	104 ± 17b	48 ± 6b	7.8 ± 0.7b

Where: CI = cardiac index (ml/min/kg); dP/dt = mm Hg/sec; EDV and ESV = LV end diastolic and systolic volumes (ml); EF = ejection fraction (%); and E_max = maximal elastance at end-systole (mm Hg/ml).

*P < 0.01 compared to corresponding control group, bp < 0.01 compared to both control and AMR groups.

These results demonstrate that LV elastance and contractile function are preserved in acute MR with intact chordae tendineae, but decline progressively in chronic MR. Furthermore, the acute MR produced in this model is well tolerated, results in progressive LV dysfunction, and is analogous to that seen in human disease.

*By Invitation

42. Metabolic Evidence for Fibroblast Injury During Cardiac Valve Harvesting, Antibiotic Disinfection and Cryopreservation: Protection with Inhibitors of Andenosine Deaminase and Adenine Nucleotide Transport

ROBERT H. MESSIER, JR., M.D.*,

ANWAR S. ABD-ELFATTAH, Ph.D.*,

RICHARD A. HOPKINS, M.D.*

PATRICK W. DOMKOWSKI, M.S.*,

DONALD G. CRESCENZO, M.D.* and

ROBERTB. WALLACE, M.D.

Washington, DC

Human cardiac valves are increasingly used in the reconstruction of ventricular outflow tracts and have been shown to offer performance advantages over porcine and mechanical prostheses; the long-term durability of these conduits is believed related to leaflet fibroblast viability at implantation. Preparation for their use as cryopreserved homografts entails multiple steps which are each potentially injurious to the leaflet cells. These include: (1) obligate harvest associated warm ischemia, (2) 24 hours of 4°C antibiotic disinfection and (3) DMSO protected cryopreservation at -1°C/min to -170°C. Using 118 semilunar porcine valves as a model of cadaveric harvest with a fixed 40 minutes of warm ischemia, leaflet cell high energy phosphates were assayed to quantify the associated metabolic injury. 58 control (Co) leaflets were processed according to the cryopreservation protocol artificially abbreviated at staggered degrees of completion to analyze stepwise catabolic damage. 60 additional leaflets were treated (Rx) at procurement with the adenine nucleotide transport inhibitor p- nitrobenzy-thioinosine and the adenosine deaminase inhibitor erythro-9-(2-hydroxy-3-nonyl) adenine to attempt nucleotide salvage of processing-incurred high energy phosphate losses. Control and treated valves of Group I were placed immediately in liquid N₂ after cardiotomy (baseline). Following harvest, Groups II and IV valves were placed for 24 hours in RPMI nutrient media (4°C) while III and V were similarly maintained with the addition of antibiotics (cefoxitin 240 µg/ml, lincomycin 120 µg/ml. polymyxin B 100 µg/ml,vancomycin 50 µg/ml). After the 24 hours, Groups II and III valves were metabolically quenched by immersion in liquid N₂ while IV and V were dimethylsulfoxide-cryopreserved at 1° C/min to - 170°C. The efficacy of nucleotide salvage was evaluated through each step by comparing High Performance Liquid Chromatography assayed ATP, ADP, and AMP and Co and Rx-treated groups. Results normalized to leaflet protein (Lowry) as nanomoles nucleotide/mg protein (± SEM).

GROUP		I (N)	II (N)	III (N)	IV (N)	V (N)	KEY: I - 40 min warm ischemia II - 1-24 hrs cold ischemia III - II with antibiotics IV = II + cryopreservation V - III + cryopreservation
ATP	Co	1.79 ± .31 (12)	1.48 ± .15 (12)	0.91 ± .19* (11)	0.46 ± .12*§ (11)	0.25 ± .10* (11)
	Rx	2.02 ± .21 (12)	1.20 ± .26 (12)	1.49 ± .11* (12)	1.58 ± .32* (12)	1.47 ± .32* (12)
TAN	Co	2.59 ± .34 (12)	2.81 ± .34 (12)	2.27 ± .50 (11)	0.67 ± .13*§ (11)	0.67 ± .13*§ (11)
	Rx	3.79 ± .36 (12)	2.84 ± .35 (12)	2.57 ± .32 (12)	2.19 ± .24* (12)	2.19 ± .24* (12)
TAN = ATP + ADP + AMP (ANOVA) * p < .05 vs Group I & II * p < .05 vs corresponding control § p < .05 vs Group III

High energy nucleotide depletion of control valves occurs independently and most significantly with antibiotic disinfection and cryopreservation (Groups I and II > III, IV, and V, p < 0.05). High energy phosphates were maintained in corresponding Rx-treated valves indicating that nucleotide salvage therapy completely protects leaflet cells from this depletion during all homograft preparation steps. Such strategies may be clinically important in the long-term performance of homograft cardiac valve transplants.

10:20 a.m. INTERMISSION - VISIT EXHIBITS

*By invitation

11:05 a.m. SIMULTANEOUS SCIENTIFIC SESSION D - ADULT CARDIAC SURGERY - Los Angeles Ballroom

43. Dynamic Cardiomyoplasty: A Seven Year Clinical Experience

ALAIN F. CARPENTIER, M.D.,

JUAN-CARLOS CHACHQUES, M.D.*,

PIERRE A. GRAND JEAN, M.S.*,

JOHN Y.M. RELLAND, M.D.*, CHRISTOPHE A CAR, M.D.*

and DANIELLE BENSASSON, M.D.*

Paris, France

Since January 1985, the date of the first dynamic cardiomyoplasty, until October 1991, 45 patients (pts) (38 males and 7 females), aged 15 to 68 years (mean 50 years) were operated upon in our Institution. The mean pre-operative NYHA functional class was 3.3 and the mean ventricular ejection fraction (EF) was 17 ± 3%at rest. Mean cardiothoracic ratio (CTR) was 54 ± 5%. Maximal oxygen consumption was 13.7 ± 3.5 ml/min/kg. The number of hospitalizations due to congestive heart failure the year prior to surgery was 2.4 hosp./pt/year. Etiology of cardiac failure was ischemic cardiomyopathy (29), idiopathic dilated cardiomyopathy (13), cardiac tumor (2), congenital post-Fontan (1). Ventricular failure involved the LV (31), the RV (3), or both ventricles (10). In 40 pts, the cardiomyoplasty was the first cardiac operation, while 5 pts had a previous operation. Surgical techniques consisted of ventricular reinforcement (32), ventricular substitution (11) and atrial reinforcement (1). The left latissimus dorsi muscle (LDM) was wrapped around the ventricles in a clockwise fashion in 41 pts, and in a counterclockwise fashion in 3 pts. Associated procedures in 20 pts comprised LV aneurysm resection (9), valve surgery (7), cABG (6), tumor resection (2). Thirteen pts were assisted perioperatively with an IABP and one with a VAD.

Pre-assist mortality rate before full LDM stimulation was 7/13 pts (54%) in the 1985-1987 period and 4/32 (12.5%) in the 1988-1991 period. The cause of death were heart failure (4), multiorgan failure (3), septicemia (2), ventricular fibrillation (1), sudden death (1). Multivariate analysis of factors influencing hospital mortality showed that age, cardiac suture technique, associated surgical procedure, biventricular heart failure, and pts hemodynamically unstable on inotropic drug support were predictors of unfavorable outcome.

All pts were followed from 3 months to 6.5 years (mean 21 months). Post-assist mortality rate was 7/34 (20.5%). Causes of deaths included heart failure (4), ventricular fibrillation (1), myocardial infarction (1), gastricbleeding (1). Pre-operative risk factors influencing long-term mortality are NYHA functional class IV, biventricular heart failure, atrial fibrillation, CTR>60%,EF< 15%.

Actuarial survival at 6 years was 11%(pre assist mortality excluded). Surviving pts were in mean NYHA functional class 1.8 (preop 3.3, p < 0.05). The average EF (rest/stress) were 25/28% at 1 years, 26/30% at 2 years, 23/28% at 3 years. Average CTR ratio were 57 ± 3% at 1 years, 56 ± 2% at 2 years, 57 ± 2.5% at 3 years. Catheterization obtained in 20 pts showed no significant changes at rest of capillary wedge pressure, pulmonary artery pressure, and diastolic left ventricular pressure at rest when compared to preoperative pressures. Average EF increased from 24 to 30.6%. Maximal oxygen consumption increased from 13.5 ± 3.5 to 17.5 ± 3 ml/min/kg. The number of rehospitalizations due to congestive heart failure was reduced to 0.4 hosp./pt/year (preop: 2.4, p < 0.05). In 62% of the pts, pharmacological therapy was reduced after surgery. Three pts required orthotopic heart transplantation 6 months, 4 years and 5 years after cardiomyoplasty. All are alive and well (mean follow-up 6 months).

Conclusions: 1) Continuous non fatiguable LDM contraction at heart rate has been obtained in the human for periods up to 6.5 years. 2) Cardiomyoplasty effectively improves the functional capacity and the survival of patients suffering from severe chronic heart failure. 3) It stops the process of continuous ventricular dilatation. 4) Hemodynamically proven benefit however remains unconstant although significant progress has been made as a result of better patient selection, new surgical techniques and better perioperative management.

*By invitation

44. Acceleration of Neointima Formation in Vascular Prostheses by Transplantation of Autologous Venous Tissue Fragments: Applicability to Small Diameter Grafts

YASUHARUNOISHIKI, M.D.*, YASUKO TOMIZAWA, M.D.*,

YOSHIHISA YAMANE, Vet, Ph.D.*,

SHINICHISATOH, M.D.*, TAKOFUMI OKOSHI, M.D.*

and AKIHIKO MA TSUMOTO, M.D.*

Yokohama, Japan

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

We developed a method to accelerate neointima formation in synthetic vascular prostheses by transplantion of autologous venous tissue fragments. A canine jugular vein was resected, minced into tissue fragments, and suspended. This was sieved through the wall of a highly porous fabric vascular prosthesis (MICROKNIT: Water porosity:4,000 ml/cm²) by pressurized injection causing tissue fragments to be trapped in the graft wall. Tissue fragmented (TF) grafts (7 mm ID, 5.7 cm long) were implanted into the thoracic aorta of 35 dogs. Small TF-grafts (4 mm ID, 4 cm long) were pretreated with heparin and implanted into both carotid arteries in 16 dogs (32 grafts). Preclotted grafts without TF were implanted into the thoracic aorta (25 dogs) and carotid arteries (6 dogs, 12 grafts) as controls. Grafts were explanted from one hr to 495 days after implantation. In the TF-grafts, host cells migrated and proliferated actively into the fibrin inner capsule. A single layer of endothelial cells formed on the luminal surface, covering multiple layers of smooth muscle cells. New arterial wall formation was complete throughout the TF-grafts within 2 weeks; however, in the control grafts, neointima formation was limited just to the anastomotic sites even after 2 months. Twenty small TF-grafts in the carotid position were patent, but all control grafts were occluded within one week. These results demonstrate that neointima formation in dogs can be enhanced in synthetic fabric prostheses; furthermore, long-term patency of small-caliber vascular grafts is possible using this tissue fragmentation technique in dogs.

*By Invitation

45. The Importance of Exploration of the Aortic Arch in Type "A" Dissection

TIRONE E. DAVID, M.D., JOANNE BOS, R.N.*

and ZHAO SUN, MS*

Toronto, Ontario

Since 1986, we have systematically explored the aortic arch of patients (pts) with type A dissection when the false lumen extends into the arch. Intimal tears are repaired or the arch is replaced; the distal anastomosis between the Dacron graft and arch or ascending aorta is performed under circulatory arrest, and cardiopulmonary bypass is re-established using antegrade arterial perfusion by cannulating the aortic arch through the Dacron graft.

From May 1981 to June 1991, 55 pts with type A aortic dissection were operated on. There were 41 men and 14 women whose mean age was 56 years, range 21 to 76. Twenty-four pts (Group I) were operated on before 1986 and 31 pts (Group II) were operated on after the new surgical approach was introduced. Preoperatively, 2 pts from Group I and 3 pts from Group II were known to have an intimal tear in the aortic arch. The false lumen involved the aortic arch in 19 pts (79%) in Group I and in 26 pts (84%) in Group II. The aortic arch was repaired or replaced in only 2 pts (8%) in Group I, and in 15 pts (48%) in Group II. There were 2 operative deaths, one in each group. Both deaths occurred in pts with extensive false lumens. The operative morbidity was similar in both groups. Contrast-enhanced CAT scans were performed in all operative survivors. Persistent false lumen was detected in 12 pts (66%) in Group I and,in only 1 pt (4%) in Group II. During a mean follow-up of 46 months further aortic surgery was necessary in 4 pts; all 4 had persistent false lumen. There were 9 late deaths; 6 were cardiovascular-related. Five of these 6 deaths occurred in pts with persistent false lumen. The overall actuarial survival at 5 years was 83% +/-5%.

Our data suggest that intimal tears in the aortic arch are common in pts with type A dissection. Repair of these tears and antegrade perfusion through the Dacron graft obliterates the false lumen in most pts, decreases the probability of further aortic surgery and may enhance survival.

12:10 p.m. ADJOURN

*By Invitation

WEDNESDAY MORNING, April 29, 1992

9:00 a.m. SIMULTANEOUS SCIENTIFIC SESSION E - GENERAL THORACIC SURGERY - Beverly Hills Room

46. Medical Tumors of the Chest Wall: Plasmacytoma and Ewing's Sarcoma

MICHAEL BURT, M.D.*, MARTINKARPEH, M.D.*,

OZURU OKOHA, M.D.* and

ROBERT J. GINSBERG, M.D.

New York, New York

Solitary plasmacytoma and Ewing's sarcoma of the chest wall are relatively uncommon tumors and data concerning treatment and results are sparse. In order to assess the results of therapy we reviewed over 40 year experience.

Methods: Records of 24 patients with solitary plasmacytoma and 62 with Ewing's sarcoma arising in chest wall admitted to our institution from 1949 to 1989 were reviewed. Survival was calculated by Kaplan-Meier method; comparisons by log rank analysis; significance defined as p < 0.05.

Results: Plasmacytoma (n = 24): Age: 35-75 yr (median 59); M:F 2.4:1. Presenting complaint was pain and/or mass in 92% (22/24). Primary therapy was local only in 5 (resection in 3, radiotherapy (RT) in 2), chemotherapy in 16 (with resection in 4, RT in 11, alone in 1), and none in 3. Subsequent multiple myeloma developed in 75% (18/24). Overall 5 year survival was 38% (median: 56 mos). Age, sex, site of primary, and local therapy did not significantly impact on survival. Ewing's sarcoma (n = 62): Age 2-39 yr (median 16); M:F 1.6:1. Presenting complaint was pain and/or mass in 90% (56/62). Primary therapy was local in 32 (resection in 22, RT in 7, resection + RT in 3) and chemotherapy in 30 (with resection in 19, RT in 5, alone in 6). Overall 5 year survival was 48% (median: 57 mos). Age, sex, and site of primary did not significantly impact on survival. Patients who developed distant metastases (n = 48) had a significantly decreased survival (5 yr: 28%) compared to those who did not (n = 14; 5 yr: 100%).

Conclusion: Plasmacytoma of chest wall, even if solitary at presentation, should be considered a systemic disease and therapy directed as such. For Ewing's sarcoma, although resection may offer local control, because of the high incidence of distant metastases (77%), systemic therapy should be considered an integral part of treatment.

*By Invitation

47. N2 Lung Cancer: Outcome in Patients With False Negative Chest CT Scans

BENEDICT D.T. DALY, M.D., JAMES D. MUELLER, M.D.*,

L. JACK PALING, M.D.*, JAMES T. DIEHL, M.D.*,

MARK S. BANKOFF, M.D.* and

DANIEL D. KARP., M.D.*

Boston, Massachusetts

Over the past 13 years, 681 consecutive patients with lung cancer have undergone computed tomographic (CT) and surgical staging of the mediastinum. Five hundred and one had negative mediastinal lymph node staging by CT and of these, 37 had cancerous mediastinal lymph nodes at mediastinoscopy (1) or thoracotomy (36). In order to determine the consequences of missing mediastinal adenopathy in patients with false negative chest CT scans, we analyzed the survival in this group of patients according to T status, central or peripheral location of tumor, cell type, areas of mediastinum involved, and the extent of nodal involvement with tumor. Twelve patients had central tumors: T2 - 5 pts., T3 - 5 pts., and T4 - 2 pts. Twenty-five had peripheral tumors: T1 - 12 pts., T2 -12 pts., and T3 -1 pt. Two of the patients in the central group died postoperatively and only two others remain alive whereas 12 of the 25 patients in the peripheral group remain alive. Four of the 37 patients, two in each group, were not resected and all are dead. One patient in each group is alive with disease. All but two of the resected 31 survivors received postoperative adjuvant XRT (23 pts.), chemotherapy (1 pt.), or XRT and chemotherapy (5 pts.). The projected two and five year survivals (Kaplan Meir) for all patients combined was 40% and 21%; for those resected 44% and 24%; for resected central tumors 40% and 0%; and for resected peripheral tumors 47% and 47%. None of these differences are significant (P>0.10). Cell type, location or number of locations of involved nodes, and the average percentage or maximum and average size or maximum size percentage of mediastinal nodes involved with tumor did not impact on survival. These data suggest that definitive thoracotomy with appropriate lymph node sampling or dissections is justified in all patients with negative chest CT scans since even patients with unsuspected N2 tumors can usually undergo resection with satisfactory 2 year survival in all patients and prolonged survival in patients with peripheral cancers.

*By Invitation

48. Comparison of Thoracoscopic Talc Pleurodesis With Standard Chest Tube Using Tetracycline and Bleomycin for Control of Malignant Pleural Effusion

DANIEL L. HARTMAN, M.D.*, PRAVEEN N. MATHUR, M.D.*,

JAMES M. GAITHER, M.D.*, DEBBIE MYLET, R.N.*,

KENNETH A KESLER, M.D.* and

JOHN W. BROWN, M.D.

Indianapolis, Indiana

Adequate treatment of malignant pleural effusions requires repeated aspirations of pleural fluid or drainage and sclerosis to relieve the distressing symptoms of breathlessness. We prospectively studied the use of in-tracavitary Talc insufflated by a Thoracoscope and compared this to historical controls; patients who participated in a randomized controlled study in which standard chest tube drainage with either Bleomycin or Tetracycline was instilled. The 25 patients in the Talc group underwent Thoracoscopy with local anesthesia and intravenous sedation. There ages ranged from 41 to 88 yrs. We recorded clinical characteristics, laboratory data, findings, and any complications associated with the procedure. Of the evaluable patients in the Talc group, 95% had a successful pleurodesis at 30 days, 92%at 60 days and at 88% at 90 days. In comparison, the bleomycin group had successful pleurodesis of 64% at 30 days and 70% at 90 days; the tetracycline group had successful pleurodesis of 33% at 30 days and 47% at 90 days. In the Talc group, one patient had extraluminal compression of the right lower lobe bronchus preventing lung reexpansion and subsequent pleurodesis. Another patient in Talc group had severe unexplained chest pain and required intensive care monitoring. In conclusion, thoracoscopically administered intracavitary Talc is a safe procedure and superior to tetracycline and bleomycin in the control of malignant pleural effusions.

*By Invitation

49. Successful Surgical Resection of Both Stages IIIA and IIIB Non-Small Cell Lung Cancer After Intensive Preoperative Chemoradiotherapy: A Southwest Oncology Group Trial

VALERIE W. RUSCH, M.D.*, KATHYS. ALBAIN, M.D.*,

JOHNCROWLEY, Ph.D.*, THOMAS RICE, M.D.*,

ROBERT LIVINGSTON, M.D.* and

JOHN R. BENFIELD, M.D.

New York, New York; Sacramento, California;

Seattle, Washington and Cleveland, Ohio

Recent studies suggest that preoperative chemo ± radiotherapy can improve the historically poor resectability and survival rate of Stage IIIA non-small cell lung cancer (NSCLC), but sometimes with significant associated morbidity/mortality. Such treatment has not been studied in IIIB NSCLC - usually considered unresectable. This Phase II trial tested the feasibility of intensive preoperative chemoradiotherapy for both IIIA and IIIB NSCLC in a multi-institutional setting that included university and community hospitals.

Methods: (1) patient (pts) with pathologically documented Tl-4 N2-3 (except pleural effusion) were eligible; (2) induction therapy was cisplatin 50 mg/m2days 1,8,29,36 + VP-16 50 mg/m² days 1-5, 29-33 + concurrent RT (4500 cGy, 180 cGy/fx); (3) resection was attempted 3-5 weeks after induction if the response was stable, partial, or complete; (4) complete nodal mapping at thoracotomy was required. Results: 122 pts entered. Complete clinical data currently available on 75 pts: median age = 58 (32-75); M:F = 49:26. 68/75 (91%) pts were eligible for surgery; 63/75 (84%) pts underwent thoracotomy; 55/75 (73%) pts, including 12 pts with "stable" response preop, had a complete resection; 4/63 (6%) pts died postoperatively.

	elig. for resection	complete resection	"complex"* resection	mean op time (hrs)	mean bid. loss (ml)	mean # days in hospital
IIIA	40	35 (88%)	13/35 (37%)	3.2 ± 1.36	741 ± 564	11.3 ± 8.7
IIIB	28	20 (72%)	6/20 (30%)	3.2 ± 1.64	547 ± 352	8.84 ± 4.53

*e.g. intrapericardial pneumonectomy, lobe + chest wall or vertebral body resection.

Complete pathology data currently available on 53 pts: 11 (21%) had no residual tumor; 20 (30%) pts had only rare microscopic foci of tumor. 2 year survival: 30% for IIIA and IIIB.

Conclusions: This intensive therapy is well tolerated; it leads to a better resectability rate and pathologic CR than most other reported regimens. Contrary to usual resection criteria this approach is applicable even to IIIB NSCLC. Surgical risk is similar to that for early stage lung cancer, even though extensive operations are often necessary. A planned randomized trial using this regimen will further assess the impact of resection on survival in IIIA/IIIB NSCLC.

10:20 a.m. INTERMISSION - VISIT EXHIBITS

*By Invitation

11:05 a.m. SIMULTANEOUS SCIENTIFIC SESSION E -GENERAL THORACIC SURGERY - Beverly Hills Room

50. Indications, Risks and Results of Completion Pneumonectomy

JOCEL YN GREGOIRE, M.D.* and

JEAN DesLAURIERS, M.D.

St. Foy, Quebec

Completion pneumonectomy refers to an operation intended to remove what is left of a lung partially resected during previous surgery. The procedure is seldom indicated and, based on current literature, it carries a significantly higher risk of operative mortality/morbidity than that of standard pneumonectomy. Over the past 20 years, 60 consecutive patients aged 17 to 70 (avg: 49.7 years) and initially diagnosed as having lung cancer (N = 43), or benign pleuro-pulmonary disease (N = 17) underwent completion pneumonectomy (Table). The interval between the first operation and completion pneumonectomy averaged 31 months for carcinoma patients and 215 months for patients with benign disease.