WEDNESDAY MORNING, MAY 5, 1982

8:30 a.m. Forum Session - Assembly Hall

32. What is the Relationship Between Plasma Heparin Concentration and (ACT) Activated Clotting Time?

RICK A. ESPOSITO*, ALFRED T. CULLIFORD*,

STEPHEN B. COLVIN*, STEPHEN T. THOMAS*,

SANFORD N. GITEL* and FRANK C. SPENCER

New York, New York

The ACT has gained popularity as a means of assessing the adequacy of anticoagulation during cardiopulmonary bypass (CPB). However, neither the precise correlation of the ACT to the plasma heparin level nor the desired therapeutic level are known. We investigated these relationships in 40 patients by measuring serial ACTs, plasma heparin levels, antithrombin III, and routine coagulation parameters. Before CPB dose-response curves were constructed with serial doses of 2 and 4 mg/kg of heparin. CPB was started when the ACT > 400 seconds, which occured in only 47% of patients after 4 mg/kg. The ACT-heparin relationhip varied to an astonishing degree (2 mg/kg: ACT 139.8 ± 43 sec, 4 mg/kg: 276 ± 91.5 sec). Heparin dosage compared to heparin concentration showed a separate wide variation, (2 mg/kg: 3.72 ± 0.96 u/ml, 4 mg/kg: 6.23 ± 1.08 u/ml). The ACT was not predictive of the heparin level at 2 mg/kg (r = 0.447 or at 4 mg/kg (r = 0.33). Heparin was neutralized with 2 mg protamine/kg, regardless of the total heparin dose with plasma heparin levels falling from 4.33 ± 1.47 to 0.20 ± .21 u/ml. If the ACT remained elevated after the 2 mg/kg dose, an additional protamine dose was calculated from the dose response curve. 49% of patients required additional protamine, but rarely as much as 3 mg/kg. Heparin levels were insignificant (<0.2 u/ml) for four hours after protamine. Total postoperative blood loss was not significantly different from a retrospective control group (528.9 ± 341.3 vs 516.7 ± 251.3cc p>.05). Hence a protamine dose of 30 to 50% less than that commonly used was effective. Heparin rebound occurred in two patients at one and two hours (heparin levels 0.57 u/ml and 0.89 u/ml), unassociated with additional blood loss. The sensitivity and specificity of the ACT to detect a plasma heparin level greater than 0.2 u/ml was 95% and 84% respectively in the postneutralization period. However the ACT did not correlate with the heparin levels at levels above 0.2 u/ml (r = 0.354).

These data show (1) ACT has no linear relationship to plasma heparin, (2) heparin sensitivity varied widely among patients, (3) 2-3 mg of protamine per kg effectively reversed heparin in almost all patients, (4) heparin rebound was not clinically significant. Whether the ACT level (greater than 400 sec) or heparin concentration (greater than 3 u/ml is the best measurement of safe anticoagulation during CPB remains uncertain.

*By invitation

33. Right Ventricular Function During Left Heart Bypass

ALFONSO T. MIYAMOTO*, SHIGEO TANAKA*

and JACK M. MATLOFF Los Angeles, California

Right heart failure has been recognized as a complication of left heart support bypass (LHBP), particularly with the most effective left ventricular transapical cannulation (TaLV) using pulsatile systems. Eight dogs (26.9 ± 1.4 kgm) were subjected to roller pump LHBP using no reservoirs in the circuit. Two inlet cannulae were connected in parallel (LA septal suture cannulation with Litwak-Koffsky LHBP cannula of 6.3 mmID, and TaLV cannulation with No. 30 venous cannula of 6.7 mmID) to allow comparison of both techniques. The cardiac output (CO) was controlled (100 ml/kg/min) by a roller pump returning the entire venous blood to the RA after disconnecting both cavae from the RA. Right ventricular function was evaluated by peak RV pressure, its first derivative (dp/dt) and mean RAP mesurements. LAP, LVP, AoP and both roller pump flows (F) were determined. Results: mean ± SEM of duplicate determinations obtained after at least 5 minutes of hemodynamic stabilization at each particular condition. (* = p<0.05 vs control; ** = p<0.05 vs 60% LHBPF ratio).

	LHBPF	x 100	LAP torr	Peak LVP torr	RV dp/dt Torr/sec	RAP torr	Peak RVP torr
	CO
	%
Control	No LHBP		2.9 ± 1.2	126 ± 4.5	212 ± 17	3.3 ± 0.9	27 ± 2.9
LHBP-Scpuil(S)	60		0.7 ± 05*	109 ± 5.7*	192 ± 16*	2.3 ± 0.6	26 ± 3.l
LHBP-S	80		0.7 ± 04*	101 ± 6.3***	182 ± 17*	2.3 ± 0.5	26 ± 3.0
LHBP-S	90		0.3 ± 0.1*	99 ± 6.31***	178 ± 16***.	2.2 ± 0.5	27 ± 3. 3
LHBP-S Max.	101		0.1 ± 0.1*	90 ± 6.4***	174 ± 16***.	2.6 ± 0.6	26 ± 3.4
LHBP-Tal.V Max.	102		1.5 ± 0.7*	72 ± 11***0	200 ± 15	3. 4 ± 0.8	26 ± 3.0
LHBP-S + TaLV Max.	105		0.1 ± 0.1*	2 ± 1.7***	168 ± 13***	5.1 ± 1.1***	25 ± 3.0

The decrease of RV dp/dt and increase of RAP are related to degree of LV decompression, i.e., LHBP flow ratio. The more complete the LHBP, the more profound its detrimental effect on RV function. Practical implications: a) it is probably better to use the minimal LHBP flow ratio that is required to maintain an adequate body perfusion, rather than the highest LHBP flow ratio to provide maximal reduction of left ventricular myocardial oxygen consumption; b) if more than 80-90% LHBP flow ratio is required to maintain adequate body perfusion, cannulation for biventricular bypass or venoarterial bypass with extracoporeal membrane oxygenation support type bypass should be made available, and c) LA septal suture cannulation provides consistently any LHBP flow ratio up to almost the entire LV output without the disadvantages of the TaLV cannulation (LV injury, bleeding) for LHBP with standard roller pumps.

34. Mechanical Assistance of the Pulmonary Circulation After Right Ventricular Exclusion

AART BRUTEL DE LA RIVIERE*, GEORGE HAASLER*,

JAMES R. MALM and DAVID BREGMAN

New York, New York

The Fontan procedure is often associated with elevated right sided pressures and low cardiac output during the early postoperative period. A dog model was established to test the effect of pulmonary artery counter-pulsation after atriopulmonary anastomosis.

After exclusion of the right ventricle by a purse string at the right AV orifice placed during inflow occlusion, a valved conduit was inserted between the right atrial appendage and the pulmonary artery, thereby obtaining a ciculatory pattern comparable to a Fontan procedure. Counterpulsation was achieved by inserting a cannula into the conduit distal to the valve in eight dogs, while in four, alternatively, Counterpulsation could comparably be achieved through a 10 mm low porosity prosthetic graft also connected to the conduit distal to the graft.

Twenty-four observations were made. Counterpulsation resulted in a mean increase in cardiac output of 48% (p<.0001). Right atrial pressure fell significantly with a mean drop of 4 mm Hg (p<.003) allowing for a further increase in right sided filling pressue by transfusion with a subsequent further increase in cardiac output. Left atrial pressure did not change significantly unless altered by transfusion.

Pulmonary vascular resistance, which was elevated after the institution of "Fontan physiology," decreased with Counterpulsation (mean decree 35%; p<.002).

Counterpulsation instituted through the 10 mm side arm graft gave similar results. Without Counterpulsation the circulatory status of the dog deteriorated rapidly. The use of a side arm graft connected to the conduit after a Fontan procedure affords easy clinical application of this method of circulatory support without the need for additional surgical intervention for decannulation. Based upon these data it is shown that mechanical assistance of the failing right atrium after atriopulmonary anastomosis is simple and highly effective.

*By invitation

35. Growth of the Left Ventricle in Compensatory Right Ventricular Hypertrophy

DOUGLAS F. LARSON*, JACQUE R. WOMBLE*,

JACK G. COPELAND*, ROBERTS. MAMMANA*

and DIANE H. RUSSELL*

Tucson, Arizona

Sponsored by: NORMAN E. SHUMWA Y Stanford, California

Pressure load models of right ventricular hypertrophy (RVH) were produced by infusion of a toxin, monocrotaline, which causes pulmonary artery fibrosis and increased vascular resistance, or injection of silica into the pulmonary vasculature to mechanically increase pulmonary vascular resistance. Also, in a volume overload model, maximal RVH was developed by a pulmonary artery to right atrial vascular shunt. In these models of RVH a Cordis Introduced was surgically inserted into the external jugular vein to allow the introduction of 7 Fr. Swan Ganz ThermodilutionR catheter for hemodynamic measurements and chronic blood sampling. During the 30 day study period, the hypertrophy events were characterized by measurements of the hemodynamics, circulating hormones, and blood constituents.

A consistent finding in these dog models of RVH was a concomitant increase in left ventricular mass. Hypertrophy in this study was defined as an increased ventricular dry weight to body weight ratio. The monocrotaline toxin induced hypertrophy resulted in an increased RVH 141% of control with a simultaneous increase of left ventricular mass of 117% of control. The silica injected dogs were found to have RVH 132% of control with an increased left ventricular mass of 122% of control. The volume overload model had RVH 166% of control with an increased left ventricular mass of 144% of control. The only consistent hormonal finding which occurred during the hypertrophy process in the above models was a marked elevation in circulating plasma epinephrine. In dogs with pressure overload and ablation of the epinephrine source by denervation of the adrenal medulla, the heart weight to body weight ratios were decreased 88% of control. Correlation of the RVH to plasma epinephrine levels resulted in r = 0.92. In the RVH models the degree of increase in left ventricular mass was correlated to the plasma epinephrine with an r = 0.88. These data implicate endogenous circulating epinephrine as a specific hormone regulating compensatory RVH. This hormone is promoting the growth of the left ventricle simultaneously with the compensatory right ventricular growth.

*By invitation

36. Fibrin Adhesive - An Important Hemostatic Adjunct in Cardiovascular Surgery

HANS GEORG BORST*, AXEL HAVERICK*,

GERD WALTERBUSCH* and WINFRIED MAATZ*,

Sponsored by: BRUNO MESSMER, Aachen, West Germany

Fibrin adhesive is a commercially available human fibrinogen cryoprecipi-tate activated by bovin thrombin and calcium. During the last 3 years this principle has been applied experimentally and clinically in our unit and was shown highly effective in controlling diffuse or localized hemorrhage from the heart and great vessels as well as in presealing prosthetic fabric. In 10 pigs bleeding from coronary arterial and venous anastomoses as well as from epicardial lacerations was effectively controlled. Water porosity of various woven and knitted vascular prostheses was reduced to zero by fibrin glue in ex vivo experiments and complete sealing was maintained in 12 grafts subsequently implanted into fully heparinized chronic dogs. Fibrin glue controlled or prevented hemorrhage in 94.5% of 236 clinical applications involving open heart surgery (190 pts) or systemic heparinization (46 pts) (Table). Neither recurrent bleeding nor complications of gluing were observed. Conclusions: Fibrin sealing of puncture holes, epi-/myocardial bleeding sites, fabric patches and high porosity vascular prostheses appears an expedient method for avoiding the hazards of continuing hemorrhage. Pregluing of prostheses eliminates the necessity of natural preclotting and knitted grafts may be used in situations otherwise requiring woven material.

	no. cases	failures %
High pressure suture lines, anastomoses, patches	107	9.3
Low pressure suture lines, lacerations; epicardial abrasions	55	2.0
Presealed woven grafts	28	3.6
Presealed knitted grafts*	46	-
	236	5.5
*heparinization only

*By invitation

37. Long Term Evaluation of Pericardia! Substitutes

PAUL J. MEUS*, JORGE A. WERNLY*,

CHARLES D. CAMPBELL*, YOSHINORI TAKANASHI*

and ROBERT L. REPLOGLE

Chicago, Illinois and Tokyo, Japan

The development of postoperative pericardia! adhesions increases the risk of cardiac reoperations because of the danger of damaging the heart, vessels or grafts. Several pericardial substitutes have been tested in the past in an attempt to facilitate reoperation with inconclusive results. This study evaluated eight different materials as pericardial substitutes. In 32 mongrels a 10 x 5 cm piece of pericardium was excised through a right thoracotomy and the defect closed with a patch. Each material tested was implanted in 4 dogs that were sacrificed at 3, 6, 9 and 12 months. At autopsy the development of adhesions and the epicardial reaction were graded as none, minimal, moderate, and severe. Histological studies of the patch, the epicardium, and the suture line were performed. The Table below lists the materials evaluated in this study and summarizes the results obtained.

	Polytetrafluoroethylene (PTFE)			Flourinated Ethylene Propylene Film	Poly-ethylene Film	Silicone Coated Polyester Fabric	Bovine Pericardium Gluteraldehyde Fixed
	Silicone Filled Film	Low Porosity Film	High Porosity Film
							Formaldehyde	Ethanol
							Preserved	Preserved
Pleural Adhesions	Minimal	Moderate	Minimal	Moderate	Moderate	None	Minimal	Minimal
Pericardial Adhesions	Minimal	Moderate	None	Severe	Severe	None	Minimal	Minimal
Epicardial Reaction	Severe	Moderate	Severe	Severe	Severe	Severe	None	None
Patch	Intact	Thickened	Incact	Torn	Torn	Incact	Incact	Intact

Our results suggest that both types of bovine pericardium were an excellent substitute. Although there was development of minimal adhesions, these were easily dissected. The underlying anatomy was clearly recognizable due to the lack of epicardial reaction. Silicone rubber coated polyester fabric was an acceptable material for the prevention of adhesions but a severe fibrous epicardial reaction impeded the recognition of the coronary arteries. Both silicone filled and high porosity PTFE films reduced adhesions but caused a severe epicardial reaction. The other synthetic materials were considered inferior due to severe epicardial reactions and/or structural deterioration.

*By invitation

38. Elective Prolongation of Atrioventricular Conduction by Multiple Discrete Cryolesions: A New Technique for the Treatment of Paroxysmal Supraventricular Tachycardia

WILLIAM L. HOLMAN*, MASA TOSHIIKESHITA*,

PETER K. SMITH*, JAMES M. DOUGLAS, JR.*,

T. BRUCE FERGUSON, JR.* and JAMES L. COX*

Durham, North Carolina

Sponsored by: DAVID C. SABISTON, JR.,

Durham, North Carolina

The most common etiology of paroxysmal Supraventricular tachycardia (PSVT) is re-entry within the A-V node. Heretofore, the only surgical treatment employed for medically refractory PSVT has been His bundle (HB) interruption, necessitating a permanent ventricular pacemaking system. The present study was designed to develop a technique for altering the input pathways of the A-V node electively in hopes of achieving permanent prolongation of A-V conduction and ablation or modification of A-V node reentrant arrhythmias. Bipolar atrial and ventricular pacing and sensing electrodes and a tri-electrode catheter positioned in the non-coronary cusp of the aorta were used to measure the pace-artifact to atrial depolarization (PA), atrial-His (AH), and His-ventricular (HV) intervals in 27 dogs. In Group I animals (n = 17), during cardiopulmonary bypass (CPB), nine separate 4-mm cryolesions (-60°C for 2 minutes) were placed around the A-V node after the position of the HB had been identified by endocardia! mapping. PA, AH, & HV intervals were measured before and at 15, 30, 60, and 180 minutes following CPB. Group II animals (N = 10) underwent identical procedures, omitting the cryolesions. Conduction times in msec: (M ± SEM)

		Pre-CPB (Control)	15 Min	Post-CPB 30 Min	1 Hour	3 Hours
GROUP I	PA	30 ± 2	28 ± 2	29 ± 2	31 ± 2	27 ± 2
	AH	73 ± 3	117 ± 11*	108 ± 10*	102 ± 6*	101 ± 8*
	HV	29 ± 1	29 ± 1	32 ± 2	31 ± 2	28 ± 1
GROUP II	PA	28 ± 2	30 ± 1	27 ± 2	28 ± 2	27 ± 2
	AH	74 ± 5	66 ± 4 *	70 ± 5	71 ± 5	67 ± 4
	HV	29 ± 1	25 ± 1	31 ± 1	31 ± 1	28 ± 2
*p less than 0.05 compared to same interval at control

Three Group I animals survived 10 weeks and all demonstrated persistent prolongation of the A-H interval. Multiple, precisely placed small cryolesions can reliably produce permanent prolongation of normal A-V conduction by altering the input pathways and conduction characteristics of the A-V node. This results in either a modification of the ventricular response to PSVT or ablation of the A-V node reentry responsible for the PSVT without necessitating an artificial ventricular pacemaker.

*By invitation

39. Prosthetic Replacements for the Thoracic Vena Cava: An Experimental Study

ANDREW C. FIORE*, JOHN W. BROWN*,

ROBERTS. CROMARTIE*, LOUIS C. OFSTEIN*,

PAMELA S. PEIGH*, NICHOLAS S. SEARS*,

WILLIAM P. DESCHNER* and HAROLD KING

Indianapolis, Indiana; Ormand Beach, Florida and

Sioux Falls, South Dakota

The ideal substitute for the thoracic vena cava continues to be a problem. Failure of an adequate prosthesis may be due in part to decreased flow, variable intrathoracic pressure and external compression by adjacent structures. Dacron (D) grafts in the venous system have a low patency rate, PTFE (Cortex), externally stented PTFE (1MPRA) and glutaraldehyde preserved porcine pericardium (GPPP) may offer alternatives, whose use in the thoracic vena cava has not been thoroughly evaluated. The purpose of the present study was to assess the short term patency of D, Cortex, IMPRA and GPPP when compared to autologous vein (AV) as a thoracic vena cava prosthesis.

Under general anesthesia, 40 adult mongrel dogs underwent right thoracotomies and the entire intrathoracic superior or inferior vena cava of each animal was replaced with a standard segment (4.5 cm (1) x 3.5 cm (w)) of knitted D (8 dogs), Cortex (8 dogs), IMPRA (12 dogs) and GPPP (12 dogs). An additional 6 animals had the same caval segment replaced with AV, fashioned as a panel graft using the external jugular vein and served as controls. After 30 days, patency was assessed by contrast venography and the implanted material removed for histological evaluation. The patency rate of each graft was compared to that of AV and the results are shown below:

Prosthesis	Patency at 30 Days	vs. Autologous Vein
AV	6/6	-
D	0/8	*P<.001
PTFE	6/8	NSD
IMPRA	12/12	NSD
GPPP	6/12	*P<.05
*Obtained by chi square analysis.

No D grafts and only 50% of GPPP grafts were patent at 30 days. Three-fourths of the Cortex grafts and all of the IMPRA prostheses remained patent.

This study demonstrated: (1) D grafts remain inadequate as venous conduits. (2) PTFE has been shown to offer higher expectations of patency when used as a thoracic venous prosthesis, than do grafts fashioned from porcine xenograft. (3) In the context of this experiment, the early patency of externally stented PTFE equals that of autologous vein in the thoracic vena cava.

9:45 a.m. Intermission - Visit Exhibits

*By invitation

10:30 a.m. Scientific Session - Assembly Hall

40. En Bloc Resection for Neoplasms of the Esophagus and Cardia

DAVID B. SKINNER

Chicago, Illinois

In 1963, Logan reported experiences with en bloc resection of carcinoma of the cardia. In 1965, a technique was developed for en bloc resection of carcinoma of the body of the esophagus with removal of the entire posterior mediastinum including thoracic duct and azygos vein system. Beginning in 1969, a radical en bloc resection for carcinoma of the cardia and esophagus was adopted for all operable cases. In 1974, en bloc approach including radical neck dissection was adopted for carcinoma of the cevical esophagus.

From June, 1969 to July, 1981, 175 patients with neoplasms of the esophagus and cardia were referred to me for treatment. Among these, 80 were considered operable based upon preoperative and intraoperative evaluation indicating that all grossly detectable disease could be encompassed by radical resection. Another 15 had palliative resection, 19 had a bypass operation, 57 were radiated for palliation after exploration or positive node or liver biopsy, and 4 had a tube inserted for palliation.

Among the 80 radical resections, there were 9 (11%) hospital deaths within 30 days of resection. Abolute survival rate for all patients operated more than 3 years ago is 22%, and actuarial table 3 year survival for the entire series is 24%. There have been no deaths for recurrent cancer after 3 years. Results were significantly worse among 12 patients receiving preoperative radiation therapy than in those 68 in which surgery was the first therapy (1 year survival 18% vs. 66% with no difference in hospital mortality). Results were similar for 39 lower third and 27 middle third tumors and slightly better for 14 cancers of the cervical esophagus. There was no difference in results between squamous and adenocarcinoma, although 18 patients with adeno-carcinoma in Barrett's esophagus had poorer results than 17 with carcinoma of the cardia.

Operative techniques, complications, rationale, and detailed results are presented.

41. Transhiatal Esophagectomy Without Thoracotomy - A Dangerous Operation?

MARK B. ORRINGER and JAY S. ORRINGER*

Ann Arbor, Michigan

In 1978, before this Association, a preliminary report describing trans-hiatal esophagectomy without thoracotomy in 26 patients was criticized for advocating a dangerous operation which violates the basic surgical principles of adequate exposure and hemostasis. This report describes our cumulative clinical experience with this operation in 134 patients: 40 with benign disease and 94 with carcinomas at various levels of the esophagus (10 pharyngeal, 20 cervicothoracic, 5 upper third, 32 middle third, and 27 distal third). Esophageal resection and reconstruction were performed in a single stage in 129 patients, and the esophageal substitute was positioned in the posterior mediastinum in the original esophageal bed in 124 patients. Continuity of the alimentary tract was restored by anastomosing the pharynx or cervical esophagus either to stomach (119 patients) or to a colonic graft (10 patients).

There have been 11 postoperative deaths (8.2%operative morality) due to myocardial infarction (3), pneumonia (3), innominate artery rupture (2), pulmonary embolus (1), and mediastinal (1) or retroperitoneal (1) infection. None was the direct result of the technique of esophagectomy. Complications included intraoperative pneumothorax (67), transient hoarseness (35), anastomotic leak (19), chylothorax (4), and tracheal laceration (2). Average intraoperative blood loss for the entire group was 1200 ml, 1100 ml for those with benign disease, 1800 ml for those with carcinoma requiring concom-mitant laryngectomy, and 900 ml for those with carcinoma undergoing esophagectomy without laryngectomy.

Of 61 patients with carcinoma who underwent esophagectomy without concommitant laryngectomy and were discharged from the hospital alive, 45 (74%) left within 14 days of operation, and another 7 (11%) left between 15 and 21 days; thus 52/61 (85%) were discharged within 3 weeks of their operation.

These data support (he contention that a thoracic incision is seldom required to resect the esophagus for either benign or malignant disease. Transhiatal esophagectomy without thoracotomy is a safe well-lolerated operation, the "hazards" of which can be minimized by careful technique and experience.

*By invitation

42. Total Fundoplication Gastroplasty (T.F.G.) - Long-Term Follow-Up in 500 Patients

ROBERT D. HENDERSON and GARY MARRYATT*

Toronto, Ontario

Five hundred patients have been treated surgically by T.F.G. for reflux control. Patients were selected because of intractable symptoms; 182 had a previous surgery; 8 scleroderma; and 31 had a peptic stricture. Surgically, a 5 cm gastroplaty tube was made over a #60 Fr bougie and reflux control achieved using total fundoplication. In the last 3 years the completion wrap length was reduced from 3 to 1.5 cm.

There was no mortality and major morbidity was 9 (1.8%) including 5 fisculae (1%). Follow-up ranges from 1.25 to 6 years; clinical 98%; radiologic 91%; manometric and pH studies 70.8%. Four anatomic recurrences are present (0.8%), one asymptomatic and three treated surgically. Seven required revision surgery (1.4%) with wrap shortening. Revision has not been required following reduction of the completion wrap to 1.5 cm. Minor residual symptoms are present in 41 patients; 12 (2.4%) with minor gastroesophageal and 9 (1.8%) cricopharyngeal dysphagia; 5 with minor bloating and 15 (3%) with nonspecific indigestion.

Radiologically 1 (0.2%) had asymptomatic reflux. Manometrically the HPZ tone rose from 12.7 to 17.57 (38.3%). Percent DMA in the lower half of the esophagus decreased from 45.5% to 28.5%. Reflux was not demonstrated by pH studies. Asymptomatic results are present in 89.8%; 8.2% have minor residual symptoms and 2% required revision surgery and are now improved. Dysphagia was originally ocassionally produced by too long a wrap, and this problem has been minimized by reducing the wrap length.

T.F.G. has proved to be a safe operative approach capable of producing effective reflux control with minor risks of analomic recurrence.

*By invitation

43. Thymectomy in Multiple Sclerosis: Preliminary Trial

THOMAS B. FERGUSON and JOHN L. TROTTER*

St. Louis, Missouri

Myasthcnia gravis (MG) is an auto-immune disease which is known lo be influenced by the thymus gland. During the past 9 years thymectomy, utilized as the initial treatment for MG and done early in the course of the disease, has produced a 90% remission rate in the young MG patient ai our institution.

Multiple sclerosis (MS) is also thought to be a disease involving the immune system. The experience with MG cited above suggests that ihymectomy may favorably influence MS. A preliminary trial was initiated in 1976, and to date 36 patients have been entered. Proven MS patients with a clearly established pattern of progression were selected. All operations were done through a sternal-split by one surgeon. The patients have had a complete neurologic and immunologic survey before and at yearly intervals after thymeetomy. All patients are more than one year post-operative (fn 33.9 months).

Results: Comparing the trial patients one year after thymeclomy to a carefully matched group of control (no operation) MS patients: (1) pyramidal functions are significantly better (P > .01) in trial patients. Other neurologic functions show no significant difference. (2) Disability is decreased (P > .05) in trial patients with relapsing-remitting MS. (3) The number of exacerbations is significantly decreased (P > .01) in trial patients. (4) Immunologic profiles show no alteration in T- and B-cells, or in mitogen studies. Cerebrospinal fluid immunoglobulins are unchanged.

We are sufficiently encouraged by these results to continue the trial. In the future, a multi-institutional study will be required to reach definite conclusions in this unpredictable disease.

*By invitation

44. Management of Air Embolism in Blunt and Penetrating Trauma

EDWARDS. YEE*, EDWARD D. VERRIER*

and ARTHUR N. THOMAS

San Francisco, California

The charts of 54 patients treated from 1970 lo 1981 were reviewed to determine the clinical outcome after treatment of air embolism from blunt (15 patients) and penetrating (18 gunshot and 21 stabbing) thoracic injuries. The diagnosis of air embolism was confirmed by the presence of air in coronary vessels, air aspirated from the heart or a major artery, or doppler findings.

Thirty-nine patients (72%) presented to the Emergency Room in profound shock (30 patients, 56%) or cardiac arrest (9 patients, 16%) and the other 15 patients (28%) deteriorated during the first twelve hours (shock 10 and arrest 5 patients). Six patients out of 40 in shock (15%) arrested unexpectedly after intubation and administration of positive ventilation. Hemoptysis or bronchial bleeding from endotracheal tube is an early sign for air embolus (8/54, 15%).

Successful management included: (1) early thoractotomy, (2) control of bronchovenous communication by hilar crossclamping, (3) maintaining normal systolic pressures with vasopressors (56%) or aortic crosscelamping (13%), and (4) prompt correction of embolic source (lung 85%, heart 11%, cava-liver 4%), which usually requires a major lung resection. The overall survival rate is (28/54, 52%). Survival correlates with mechanism of injury (blunt - 4/15, 27% vs penetrating - 24/39, 67%, 0.10 >p>0.05) and the presence of associated nonthoracic injuries (present 9/33, 27% vs absent 19/21, 90% - p<.001).

We conclude that: (1) air embolus can insidiously occur even in blunt trauma, (2) suspicion should be high with occurrence of hemoptysis or unexpected arrest during positive ventilation, (3) treatment should include early proximal control of hilum, maintaining coronary perfusion pressures, and prompt correction of embolic sources, and (4) successful results correlate with the outcome of associated injuries.

1:00 p.m. ADJOURNMENT

*By invitation