Arthur D. Boyd

The Long-term Influence of Coronary Bypass Grafts on Myocardial Infarction and Survival

Approximately 1,000 coronary bypass procedures were performed at New York University between February 1968 and December 1973. This report reviews all elective operations performed for angina between 1968 and 1972, a total of 448 patients. In this five-year period the percentage of diseased arteries bypassed rose from 40% to 84%, and operative mortality decreased from 28% to less than 3%. There were a total of 28 operative deaths, mostly from myocardial infarction and low cardiac output. Operability was nearly 95%. The only fixed contraindication was chronic congestive failure. Over one-half of the patients had an abnormal ventriculogram, and there was some history of mild congestive failure in nearly 20%. Elevation of left ventricular end-diastolic pressure above 20 mm before operation was associated with a higher operative mortality, but the late mortality was similar to those with a normal preoperative end-diastolic pressure. In 383 surviving patients, angina was eliminated or greatly improved in 86%, unimproved in 12% and worse in 2%. Late angiograms were performed on 201 patients, studying a total of 445 venous grafts with an overall patency rate of 71%. Graft occlusion was sporadic and unpredictable, but over 90% of patients with multiple grafts remained with at least one patent graft. A late myocardial infarction occurred in 32 out of 420 patients surviving operation, and was fatal in eight. The cumulative incidence over a period of five years was 17%. Twenty-three deaths occurred following discharge from the hospital. Life-table analyses showed a five-year survival of 77% when all deaths were included, and a five-year cardiac survival of 81% when non-cardiac deaths were withdrawn alive at the time of death. The expected survival in a comparable population group without coronary disease was 92%, while data published by Sones of patients treated without operation showed a five-year cardiac survival of 66%. Current operative techniques have an operative mortality of 2-3% and a subclinical infarction rate of 5-10%. The ideal graft is yet evolving, but data with internal mammary artery grafts are most encouraging. A future goal should be a five-year graft patency of at least 80%. Because many infarcts probably develop from a relatively small decrease in coronary blood flow, either during rest or mild activity, the likelihood that future data will demonstrate a marked increase in longevity with bypass grafting is great.

Immediate Reconstruction of Full-Thickness Chest Wall Defects

Twenty-one patients had full-thickness chest wall defects reconstructed at the New York University Medical Center in the last ten years. Marlex mesh provided chest wall stability in 5 patients. In 9 patients with radiation ulcers Marlex mesh was not required; a severe fibrotic reaction had obliterated the pleural space and prevented paradoxical motion. Partial sternal resections did not require Marlex stabilization, while a total sternectomy resulted in marked ventilatory insufficiency in a patient who would have benefited from the use of a stabilizing material. Random pattern flaps were used initially; more recently, axial pattern, myocutaneous, and myocutaneous free flaps were employed. Necrosis developed in 4 (36%) of the 11 patients with random pattern flaps, but was not seen with the newer flap techniques. Myocutaneous free flaps provided uncomplicated coverage of and stability to three large, potentially contaminated defects. It seems that with the currently available flap techniques and the methods of chest wall stabilization, immediate repair of all full-thickness chest wall defects is possible.

Long-term results in 1375 patients undergoing valve replacement with the Starr-Edwards cloth-covered steel ball prosthesis.

The two principal considerations with prosthetic valves are durability and thromboembolism. With the widespread interest in recently developed prosthetic valves (porcine, tilting dise, Cooley), the long-term results at one institution with a single prosthesis were considered of particular importance. Accordingly, a 97% follow-up has been completed on 1375 patients (pts) undergoing prosthetic valve replacement with the Starr-Edwards cloth-covered steel ball prosthesis at New York University between October 1967 and December 1975. Operative procedures were as follows: aortic valve replacement (AVR): 470 pts; mitral valve replacement (MVR): 362 pts; combined AYR and MVR: 129 pts; other combined procedures: 414 pts. Overall operative deaths were 13.7%, 9% for AVR, 10.8% for MVR, and 18.6% for combined AVR and MVR. At seven years, AVR survival was 64%, and MVR survival 64.5%. There has been widespread pessimism, usually without significant data, about the cloth-covered prosthesis, because of concern of cloth wear, hemolysis and other complications. Therefore, a particularly significant finding by actuarial analysis was that 85% of surviving patients with isolated AVR remained free of emboli for five years. In pts surviving isolated MVR, 80% remained free of emboli for five years. Of those having embolie episodes, 33% were not on anticoagulants. Fatal hemorrhage from anticoagulants occurred in 0.8% of pts. Endocarditis occurred in 5.7% of the entire group, with 1.3% requiring reoperation. Clinically significant hemolysis occurred in 5.1% of the group, with only 0.2% requiring reoperation. Hence, the total frequency of clinically significant cloth-wear was less than 0.5%. These data indicate both the reliability and the limitations of the Starr-Edwards cloth-covered steel ball valve and can be used in comparing experiences with the more recently developed prostheses.

Treatment of AIDS-related bronchopleural fistula by pleurectomy☆

Spontaneous pneumothorax in patients with acquired immunodeficiency syndrome (AIDS) may require prolonged therapy for treatment of a persistent bronchopleural fistula, and treatment by standard methods often fails. This pilot study was done to test the effectiveness of aggressive surgical therapy for definitive treatment of persistent bronchopleural fistula in patients with AIDS. Between March 1989 and September 1991, 44 patients with AIDS were treated for spontaneous pneumothorax with closed tube thoracostomy; 14 of these patients had development of persistent bronchopleural fistula for more than 10 days, and 2 patients had subsequent bronchopleural fistula on the opposite side. Operative therapy in 14 patients included 15 thoracotomies and one sternotomy. The bronchopleural fistula was closed directly with suture or staples in 15 procedures and resected by lobectomy in 1 patient. All 14 patients received adjuvant parietal pleurectomy. Operative mortality was 7% (1 of 14 patients). The fistula was closed in all survivors and 13 patients were discharged between 7 and 28 days postoperatively. Pathologic examination confirmed Pneumocystis carinii in 13 patients with a high incidence of diffuse involvement and subpleural necrosis, further demonstrating the need for pleurectomy. These data suggest that in selected patients bronchopleural fistulas associated with AIDS can be effectively controlled by surgical closure combined with pleurectomy.

Myocardial injury associated with potassium arrest

The relative efficacy of potassium-induced ischemic arrest using buffered, isosmotic potassium (25 mEq/liter) was compared with hypothermic arrest in an experimental protocol employing an intact canine heart preparation. Myocardial function (LVSW, dp/dt max), serum creatine phosphokinase levels, myocardial perfusion, and light and electron microscopical examination of the heart were assessed in five groups of 5 dogs each. There was one control group (90 minutes of bypass, no anoxia) and four experimental groups, each subjected to 1 hour of ischemic arrest and 30 minutes of reperfusion, comparing normothermic ischemic arrest (NIA), hypothermic ischemic arrest (myocardial temperature less than 25 degrees C) (HIA), normothermic potassium arrest (NKA), and hypothermic potassium arrest (HKA). Myocardial function decreased significantly following NIA and NKA but remained essentially equal in the control, HIA and HKA groups. Serum creatine phosphokinase analysis documented a significant increase in each group of animals: 2,250 mU after NIA, 1,778 mU after NKA, 1,388 mU after HIA, 1,220 mU after HKA, and 838 mU after control bypass. Left ventricular myocardial perfusion was unmeasurably low after NIA, reduced to 111 m/100 gm of tissue/min after NKA, and increased to 165 to 188 ml/100 gm/min in the control, HIA and HKA groups. Electron microscopical studies showed a range of myocardial changes, from probably irreversible damage after NIA to similar but less diffuse changes after NKA, and to potentially reversible changes after HKA and HIA with the least alteration from control after HIA. The results indicate that potassium arrest alone is not as effective as hypothermia in preventing ischemic injury, and the combination of hypothermia with a single 150 cc administration of potassium (25 mEq/liter) does not appear to provide significant additional protection.

Late Complications of Intraoperative Coronary Artery Perfusion

Although exceedingly unusual, six significant late and potentially lethal complications of coronary perfusion have been observed. In five patients, angina was a sequel of aortic valve replacement, and in the sixth, ventricular tachycardia occurred. Coronary angiography revealed proximal coronary artery stenosis not present in preoperative study in five of the patients, and a dissecting aneurysm of the right coronary artery in the sixth. In four patients, the findings were confirmed during successful reoperation. These patients illustrate the urgent need for repeating coronary angiography when angina or serious arrhythmia occurs late following aortic valve replacement.

Nonthrombogenic aortic and vena caval bypass using heparin-coated tubes

Abstract In twenty-five acute canine experiments polyvinyl chloride (PVC) tubes, both uncoated and coated with tridodecylmethylammonium chloride/heparin (TDMAC/heparin) complex, were tested in thoracic aortic bypass lasting five hours and in inferior vena caval bypass lasting thirty minutes. In aortic bypass, heparin-coated tubes were almost totally nonthrombogenic in contrast to uncoated tubes which had definite thrombus formation. Smoothly tapered tubes were significantly less thrombogenic than were hybrid tubes with junctions. Tube design, especially junction-free construction, was more important than heparin coating in preventing thrombus formation. In vena caval bypass heparin-coated tubes were totally nonthrombogenic in contrast to uncoated tubes which did show thrombus formation and thrombo-embolism. The smoothly tapered TDMAC/heparin-coated shunt tubes were successfully employed in four patients during thoracic aortic aneurysmal resection.

The effect of diuretics on renal hemodynamics during cardiopulmonary bypass

Abstract The effect of cardiopulmonary bypass (CPB) on renal hemodynamics was studied in 15 dogs using 133xenon washout. Ten control dogs had no diuretics administered and five diuretic dogs were given furosemide immediately before and during CPB. A catheter was inserted into the right renal artery under fluoroscopic guidance via the left femoral artery and a bolus of 133xenon injected. Washout curves were obtained with a collimater placed over the kidney before CPB and after 15 and 90 min of CPB. Total CPB was undertaken at normothermia using venous gravity drainage, an arterial roller pump, a heat exchanger and a Kolobow membrane oxygenator. Washout curves were analyzed and four components of renal blood flow (RBF) developed: I, cortex; II, juxtamedulla; III, inner medulla; and IV, hilar fat. Percentage of total radioactivity and regional blood flow was derived for each component and total RBF calculated. Total RBF in the control group decreased progressively during CPB (457 → 269 → 158 ml/100 g/min after 90-min CPB). This decrease in RBF was associated with a marked shunt of flow from cortical to juxtamedullary region. Percentage of flow to the cortex decreased as well (68% → 33% → 23% after 90 min) while activity to the juxtamedulla increased (24% → 47% → 50% after 90 min). RBF in the diuretic treated group decreased during CPB but significantly less than in the control group (436 → 397 → 273 ml/100 g/min after 90-min CPB). The intrarenal shunt seen in the control group during CPB was significantly reduced in the diuretic treated group. While percent flow to the cortex was reduced as bypass progressed (67% → 56% → 47% after 90 min), cortical activity remained greater than juxtamedullary (29% → 36% → 38% after 90 min) throughout the 2 hr of bypass. It was apparent from this study that cortical ischemia associated with CPB can be substantially reduced by diuretic therapy during CPB.

Chevalier Jackson: The Father of American Bronchoesophagoscopy

Chevalier Jackson was involved with bronchoesophagoscopy from the late 1800s until shortly before his death in 1958. A pioneer in the field, he developed numerous instruments used in peroral endoscopy and taught their safe and effective use. Most of the next generation of leaders in the field of bronchoesophagoscopy were trained in his clinics in Pittsburgh and in Philadelphia, where he held academic appointments at the six leading medical institutions of these cities. He was a keen observer, inventor, prolific writer, and humanitarian.

The influence of diabetes and hypertension on the results of coronary revascularization

The effects of diabetes and hypertension on the early postoperative course of patients undergoing coronary revascularization were studied by reviewing the records of 177 patients operated upon in 1972. There were 121 nondiabetic, nonhypertensive; 32 hypertensive; ten diabetic; and 14 diabetic-hypertensive patients. The incidence of postoperative low cardiac output, renal insufficiency and arrhythmia was significantly higher in the hypertensive patient. Operative mortality ranged from 0 in diabetic patients, to 0.8 per cent in nondiabetic, nonhypertensives, to 7.1 per cent in diabetic-hypertensives and 12.5 per cent in hypertensive patients, suggesting an increased risk for the hypertensive patient. The one- to two-year follow-up results documented symptomatic improvement in 90.7 per cent of patients with little adverse effect apparent from diabetes or hypertension. Pre- and postoperative coronary angiography was carried out in 103 patients between 1968 and 1973 with a mean elapsed time between operation and postoperative angiogram of 9.3 months. The progression of atherosclerosis was graded on a 0-4 basis in both grafted and ungrafted coronary arteries. While hypertension appeared to contribute to disease progression, the incidence of vein graft and internal mammary artery bypass occlusion was not significantly affected by either diabetes or hypertension. This study has shown that while hypertension contributes to increased morbidity and mortality in the early postoperative period and an increased rate of progression of atherosclerosis, neither diabetes nor hypertension appeared to influence the one- to two-year results of coronary revascularization.