Transhiatal esophagectomy for benign and malignant disease.

Transhiatal esophagectomy has been performed in 583 patients with diseases of the intrathoracic esophagus: 166 (28%) benign and 417 (72%) malignant (6% upper, 28% middle, and 66% lower third and cardia). The benign esophageal diseases included strictures (40%); neuromotor dysfunction-achalasia (24%), esophageal spasm (8%); recurrent gastroesophageal reflux (16%); acute perforation (5%); acute caustic injury (2%); and others (3%). Among the patients with benign disease, 60% had undergone at least one prior esophageal operation. Transhiatal esophagectomy was possible in 97% of patients in whom it was attempted, 19 patients (13 with benign disease and 6 with carcinoma) requiring addition of a thoracotomy for esophageal resection. Esophageal resection and reconstruction were performed in a single operation in all but 5 patients. The esophageal substitute was positioned in the posterior mediastinum in the original esophageal bed in 96%. Stomach was used to replace the esophagus in 553 patients (95%) and colon in 28 (5%) who had undergone prior gastric resections. Overall hospital mortality was 5% in patients with benign disease and 5% in those with carcinoma. There was 1 intraoperative death caused by uncontrollable hemorrhage. Complications included intraoperative entry into a pleural cavity necessitating a chest tube (74%), anastomotic leak (9%), recurrent laryngeal nerve paralysis (3%), and chylothorax and tracheal laceration (< 1% each). Three patients required reoperation for mediastinal bleeding. Average intraoperative blood loss was 875 ml (1023 ml for benign disease and 817 ml for carcinoma). Of the surviving patients, 88% were discharged able to swallow within 3 weeks of operation and 78% within 2 weeks. The actuarial survival of the patients with carcinoma is similar to that reported after more traditional transthoracic esophagectomy. Among patients with benign disease, good or excellent functional results have been achieved in nearly 70% after a cervical esophagogastric anastomosis. Although approximately 44% have required one or more anastomotic dilations within 1 to 3 months of operation, true anastomotic strictures have developed in 10%. Clinically troublesome nocturnal reflux has occurred in 3%. Transhiatal esophagectomy is feasible in most patients requiring esophageal resection for either benign or malignant disease and is a safe, well-tolerated operation if performed with care and for the proper indications.

Pseudoaneurysm of pulmonary artery in mucormycosis. Radiographic characteristics and management.

Pulmonary mucormycosis is a rare and almost invariably fatal complication that can occur in the context of severe deficits in host defenses. Antemortem diagnosis is difficult and requires a high index of suspicion together with invasive diagnostic techniques. Mucor species exhibit a pronounced affinity to invade vessels; mucormycosis involving the pulmonary vasculature has rarely been documented antemortem, and survival in this context has been rare. In this report, we describe a patient with chronic renal failure and systemic lupus erythematosus who developed extensive invasion of the left main pulmonary artery by mucormycosis. Chest computed tomographic (CT) scans and pulmonary arteriogram demonstrated a massive pseudoaneurysm of the left pulmonary artery; these radiographic findings have not previously been described in mucormycosis. Aggressive combination therapy, employing preoperative amphotericin B (AmB) followed by surgical resection (pneumonectomy) and a full course of AmB, was curative. This favorable outcome supports the role of surgery as adjunctive therapy, and it underscores the need for early diagnosis and aggressive treatment.

Automatic internal cardioverter defibrillator: a bridge to heart transplantation.

The automatic internal cardioverter defibrillator (AICD) is effective in preventing death in patients with malignant ventricular arrhythmias (VT/VF) refractory to medical therapy. Because of the long waiting period for heart transplantation and the high likelihood of sudden arrhythmic death in this population, this study was undertaken to assess the value of the AICD in patients awaiting heart transplantation who have refractory VT/VF. Fourteen patients awaiting heart transplantation who had a history of VT/VF underwent AICD implantation (10 extrapericardial and four intrapericardial) via median sternotomy. All patients survived the AICD implantation and have either had heart transplantation or await transplantation at present (1 to 24 months after AICD implantation). Twelve of these patients have received a mean of 10 AICD shocks (range, 0 to 32). One patient received 19 shocks in the 24-hour period before transplantation. Two patients have died of progressive heart failure. Five patients have gone on to successful transplantation, and seven patients await heart transplantation with a functioning AICD in place. In conclusion, the AICD represents a new "bridge" to heart transplantation that is well tolerated by these high-risk patients, avoids drug side effects, and is efficacious in aborting sudden death, thereby allowing them to undergo successful heart transplantation.

Effects of ischemia on epicardial segment shortening.

To evaluate the effects of nontransmural ischemia on epicardial contractile function, we implanted sonomicrometers in 15 open-chest, anesthetized (halothane) dogs. One cylindrical crystal (radiating ultrasound 360 degrees) was used as a transmitter for three conventional flat plate crystals arrayed to measure epicardial segment shortening along three different axes that were deviated 0 degree (parallel), 45 degrees (oblique), and 90 degrees (perpendicular) from surface fiber orientation in the anteroapical or posterior-basal left ventricle. During baseline conditions, epicardial shortening was maximal parallel with fiber orientation. Shortening decreased in a non-linear manner as deviation from fiber orientation increased, but there were significant differences between the two left ventricular regions suggesting that more substantial lateral strain occurs in the anterior-apical than the posterior-basal area. During coronary inflow restriction, changes in epicardial segment shortening also varied greatly depending on location and alignment. At levels of wall thickening impairment associated with normal subepicardial perfusion, changes in epicardial function were restricted to the segments aligned perpendicular to fiber orientation whereas the parallel and oblique segments displayed moderate dysfunction or none at all. Thus, transmural tethering modifies epicardial segmental motion during coronary inflow restriction, but the severity of the influence depends on the alignment and location of the epicardial measurements.

Descending necrotizing mediastinitis: transcervical drainage is not enough.

One of the most lethal forms of mediastinitis is descending necrotizing mediastinitis, in which infection arising from the oropharynx spreads to the mediastinum. Two recently treated patients are reported, and the English-language literature on this disease is reviewed from 1960 to the present. Despite the development of computed tomographic scanning to aid in the early diagnosis of mediastinitis, the mortality for descending necrotizing mediastinitis has not changed over the past 30 years, in large part because of continued dependence on transcervical mediastinal drainage. Although transcervical drainage is usually effective in the treatment of acute mediastinitis due to a cervical esophageal perforation, this approach in the patient with descending necrotizing mediastinitis fails to provide adequate drainage and predisposes to sepsis and a poor outcome. In addition to cervical drainage, aggressive, early mediastinal exploration--debridement and drainage through a subxiphoid incision or thoracotomy--is advocated to salvage the patient with descending necrotizing mediastinitis.

Esophagectomy for esophageal disruption.

When esophageal disruption occurs in the presence of preexisting esophageal disease or is associated with sepsis or fluid and electrolyte imbalance, aggressive and definitive therapy often provides the only chance for patient salvage. Twenty-four adults (average age, 59 years) with intrathoracic esophageal perforations underwent esophagectomy: 15, transhiatal esophagectomy without thoracotomy; and 9, transthoracic esophagectomy. Restoration of alimentary continuity with an immediate cervical esophagogastric anastomosis was carried out in 13 patients. Eleven underwent a cervical or anterior thoracic esophagostomy, and 10 of them had a subsequent colonic (7) or gastric (3) interposition from 4 to 32 weeks (average time, 8.6 weeks) later. The perforations were due to esophageal instrumentation (9 patients), acute caustic ingestion (2), emesis (2), intrathoracic esophagogastric anastomotic disruption (2), and other causes (9). Preexisting esophageal disease in 20 patients included chronic strictures (10 patients), reflux esophagitis (3), esophageal cancer (3), achalasia (2), diffuse spasm (2), and monilial esophagitis (1 patient). Ten patients were operated on within 12 hours after the injury; 3, within 12 to 24 hours; and 11, within three to 45 days (average interval, 6.6 days). There were three hospital deaths (13%). Nineteen of the 21 survivors were able to swallow comfortably until the time of death or latest follow-up. Aggressive diagnosis and aggressive treatment of life-threatening esophageal perforations are advocated. Conservative procedures (repair, diversion, or drainage) for a perforation with preexisting esophageal disease often inflict more morbidity than esophageal resection, which eliminates the perforation, the source of sepsis, and the underlying esophageal disease. The decision to restore alimentary continuity in a single stage must be individualized.

Distribution of cardioplegic solution infused antegradely and retrogradely in normal canine hearts.

The adequacy of retrograde delivery of cardioplegic solution to the right ventricle and interventricular septum is controversial. To address this issue quantitatively, we infused blood cardioplegic solution labeled with radioactive microspheres (15 microns diameter) into the coronary sinus (n = 8 dogs) at a pressure of 51 +/- 1 mm Hg (mean +/- standard error of the mean) to be compared with the same quantity of labeled cardioplegic solution (20 ml/kg) delivered through the aorta (n = 6 dogs) at 97 +/- 7 mm Hg. Both methods of delivery produced cardiac arrest, but retrograde infusion required a significantly longer time to complete the infusion (6.2 +/- 0.8 minutes versus 1.5 +/- 0.1 minutes, p less than 0.01). Greater than 99% of the microspheres passing through the vasculature of the left ventricle were trapped in the left ventricular myocardium with antegrade infusion, and the distribution of the cardioplegic solution was uniform. Antegrade delivery (cardioplegic flow x infusion time) averaged approximately 3.0 to 4.0 ml/gm, except at the apex, where delivery averaged approximately 2.0 ml/gm. With retrograde infusion, 93% of the microspheres perfusing the left ventricle were trapped and delivery of the cardioplegic solution was not uniform. In the anterolateral free wall, delivery of cardioplegic solution averaged between 1.5 and 2.9 ml/gm (p less than 0.001 compared with antegrade) and only 0.6 to 0.8 ml/gm in the posteroseptal region of the basal left ventricle (p less than 0.001 compared with the antegrade group and anterolateral samples of the retrograde group). In the middle portion of the right ventricle, antegrade trapping of microspheres was 99% and delivery of cardioplegic solution averaged approximately 2.0 ml/gm. With retrograde delivery, only 16.5% (range 11.8% to 26.0%) of the microspheres passing through the right ventricular vasculature were trapped in the right ventricular myocardium, which indicates that substantial shunting had occurred. Corrected for the high shunt fraction, retrograde delivery of cardioplegic solution to the middle portion of the right ventricle averaged only 0.5 ml/gm (p less than 0.01). Retrograde delivery to the atrial septum and right atrium was also low. Because retrograde delivery of cardioplegic solution was markedly nonuniform, we conclude that inadequate cardioplegic delivery to the middle portion of the right ventricle and posteroseptal portion of the left ventricle could result with cardioplegic infusion through the coronary sinus.

F(ab')2 fragments of anti-Mo1 (904) monoclonal antibodies do not prevent myocardial stunning.

To determine if inhibition of leukocyte adhesion and aggregation could improve postischemic ventricular dysfunction ("stunning"), a monoclonal antibody (904) that binds to the adhesion-promoting Mo1 glycoprotein on the cell surface of leukocytes was administered intravenously (0.5 mg/kg) to open-chest dogs before a 15-minute coronary occlusion. Ultrasonic crystals placed in ischemic and control myocardium were used to measure systolic wall thickening during a 15-minute occlusion of the left anterior descending artery and for 3 hours after reperfusion. Myocardial blood flow was measured with tracer-labeled microspheres before occlusion, after 10 minutes of occlusion, 3 minutes of reperfusion, and at 1 and 3 hours after reperfusion. Six animals receiving anti-Mo1 antibody had antibody excess demonstrated with immunofluorescence techniques at 5 minutes and 3 hours of reperfusion; this finding indicated saturation of binding sites. Five animals served as controls and received an antibody (murine immunoglobulin G) that does not influence neutrophils. The two groups did not differ hemodynamically during ischemia and reperfusion. Risk areas and myocardial blood flow were also not significantly different between the two groups. The main parameter used to define regional myocardial stunning, percentage systolic wall thickening in the ischemic/reperfused area, did not differ significantly between the two groups. Specimens from nonischemic myocardium were compared with ischemic specimens for myeloperoxidase content. There were no significant differences within or between groups. These data indicate that the anti-Mo1 monoclonal antibody (904) is not effective in improving the profound myocardial dysfunction that persists for 3 hours of reperfusion after 15 minutes of ischemia.

Comparison of polyclonal antibody sera for early prophylaxis following cardiac transplantation.

In order to test different polyclonal antibody regimes as early prophylaxis against cardiac rejection, 42 patients (ages 30 to 60 years) transplanted at the University of Michigan from December 1986 to August 1988 were randomized to receive antithymocyte globulin (ATGAM, Upjohn, n = 19) or antilymphoblast globulin (MALG, University of Minnesota, n = 23). Cyclosporine (CYA), steroids, and azathioprine (AZA) administration was similar in all randomized patients during early prophylaxis. CYA was begun preoperatively and maintained at a serum level of 250-300 ng/ml. After an initial steroid taper, patients were maintained on 0.3 mg/kg/day. AZA was begun after polyclonal prophylaxis at 1-2 mg/kg. All patients received either ATGAM or MALG for 7 days or until the serum CYA reached 250 ng/ml. Although sex, pretransplant hemodynamics, follow-up length, total drug dose, mortality (one per group), postoperative white blood cell and lymphocyte counts did not differ between groups, MALG significantly delayed the first rejection episode as compared to ATGAM (35 +/- 4 vs 22 +/- 3 days, P less than 0.05). Additionally, there was decreased rejection during follow-up for the MALG group with 1.5 +/- 0.2 rejections per patient as compared to 2.3 +/- 0.3 with ATGAM. Furthermore, the significant infection rate with MALG was only half that of the ATGAM group (6/23 vs 11/19) (P less than 0.05). The beneficial effect of MALG may be due to immune-specific differences in its polyclonal spectrum.

Esophageal resection for achalasia: indications and results.

Although esophagomyotomy is highly effective as the initial surgical treatment of most patients with achalasia, those with either recurrent symptoms after a previous esophagomyotomy or a megaesophagus do not respond as well to esophagomyotomy. Total thoracic esophagectomy was performed in 26 patients (average age, 49 years) with achalasia. Eighteen had a history of a previous esophagomyotomy, and 18 had a megaesophagus (esophageal diameter of 8 cm or larger). In 24 patients, a transhiatal esophagectomy without thoracotomy was the operative approach; 2 patients required a transthoracic esophagectomy because of intrathoracic adhesions from prior operations. The stomach was used as the esophageal substitute in all patients; it was positioned in the posterior mediastinum, and a cervical anastomosis was performed. Intraoperative blood loss averaged 765 mL. Major postoperative complications included mediastinal bleeding requiring thoracotomy (2), chylothorax (2), and anastomotic leak (1). There were no postoperative deaths. The average postoperative hospital stay was ten days. Follow-up is complete and ranges from 3 to 91 months (average duration, 30 months). All but 1 patient with severe psychiatric disease eat a regular, unrestricted diet without postprandial regurgitation. Early postoperative anastomotic dilation was required in 10 patients. Dumping syndrome has occurred in 5 patients. It is concluded that esophagectomy provides the most reliable treatment of esophageal obstruction, pulmonary complications, and potential late development of carcinoma in the patient with a megaesophagus of achalasia or a failed prior esophagomyotomy and that it is a far better option in these patients than esophagomyotomy, cardioplasty procedures, or limited esophageal resection.

Continued assessment of the combined Collis-Nissen operation.

The combined Collis-Nissen operation has been performed in 353 patients. Forty-five percent had reflux esophagitis without stricture; 20%, peptic stricture; 72%, a sliding hiatal hernia; 17%, a paraesophageal hernia; 21%, previous antireflux operation; 15%, esophageal spasm; 8%, scleroderma; and 32%, marked obesity. There were 4 postoperative deaths (mortality rate, 1.1%). Complications occurred in 28 patients (8%) and included wound infection (2.2%), esophageal or gastroplasty tube leak (1.7%), bleeding (1.1%), splenic injury, gastric atony, and crural repair dehiscence (each less than 1%). Follow-up includes personal interview, esophageal manometry, and standard acid reflux testing. The average length of follow-up for 261 patients (74%) followed at least 12 months is 43.8 months. Fifty-eight percent have been followed at least 36 months; 41%, 48 months; and 29%, 60 months or longer. Subjectively, in these 261 patients, reflux has been eliminated in 75%, is mild in 11%, is moderate in 9%, and is severe in 5%. Eight percent have postthoracotomy pain; 3%, early satiety ("bloats"); and 1%, postvagotomy diarrhea. Seventeen percent require either periodic or regular esophageal dilations for dysphagia. Objectively, intraesophageal pH studies show good reflux control in 91% and poor reflux control in 9%. Twenty-six patients (10%) have required reoperation for recurrent reflux or dysphagia. These results substantiate satisfactory reflux control using the Collis-Nissen operation in patients at risk for recurrence after standard repairs, but also emphasize that, like other antireflux procedures, the Collis-Nissen operation is not without some degree of postoperative adverse symptoms.

Cervical esophagogastric anastomosis for benign disease. Functional results.

Ninety-one adult patients (average age 49 years) with various benign esophageal disorders treated by total thoracic esophagectomy and a cervical esophagogastric anastomosis have been followed up with personal interviews and examinations from 6 to 104 months (average 34 months). Outpatient esophageal dilation has been used liberally for any degree of postoperative cervical dysphagia. At their latest follow-up, 39 patients (43%) eat without dysphagia; four patients (4%) have mild dysphagia necessitating no treatment; 34 patients (37%) have undergone one to three dilations during the first 6 to 12 postoperative months for intermittent dysphagia; and 14 patients (16%) have more severe dysphagia necessitating regular anastomotic dilations (two thirds of these perform home self-dilations). Mild regurgitation of gastric contents has been experienced by 27 (30%), particularly when recumbent after eating, but only four patients sleep with the head of the bed elevated to prevent nocturnal regurgitation. No patient has had pulmonary complications resulting from aspiration. Twenty patients (22%) have had varying degrees of "dumping syndrome," generally transient and well controlled with medication. Two patients have required an additional gastric drainage operation 16 months and 82 months, respectively, after the esophagectomy. At their latest evaluation, 33% of the patients weigh 3 to 83 (average 19) pounds more than they weighed preoperatively, 38% weigh 5 to 40 (average 12) pounds less, and 29% have had no change in their weight. The stomach functions well as a visceral esophageal substitute and, like the esophagus, is more thick-walled and resilient than colon. Significant gastroesophageal reflux is uncommon after a properly performed cervical esophagogastric anastomosis. Postoperative dysphagia can be minimized by attention to technique in constructing the anastomosis. These data support our belief that the stomach is the preferred organ for esophageal replacement, not only for carcinoma, but also for benign diseases as well.

The combined Collis-Nissen operation for esophageal reflux strictures.

This report evaluates the efficacy of the combined Collis-Nissen operation in achieving long-term reflux control in patients with reflux strictures. A Collis-Nissen procedure with dilation of a reflux stricture was performed in 64 adults. The strictures were mild (easily dilated) in 37, moderate (requiring some force to dilate) in 17, and severe (requiring very forceful dilation) in 10. Two strictures were perforated intraoperatively. There was 1 postoperative death, and 4 patients have been lost to follow-up. The remaining 59 patients have been followed from 2 to 120 months (average, 43 months) after operation. Subjectively, reflux is absent in 48 (81%), mild in 4 (7%), moderate in 5 (9%), and severe in 2 (3%). Objectively, intraesophageal pH studies show good or excellent reflux control in 94% at 1 year and 66% at 2 to 5 years. Dysphagia has been eliminated in 71%, is mild in 10%, moderate (requiring occasional dilation) in 12%, and severe (requiring regular dilations) in 7%. The combined Collis-Nissen operation provides good long-term reflux control and relief of dysphagia in most patients with reflux strictures. However, patients with reflux strictures after previous repairs are likely to have unsatisfactory results and may best be managed with resectional therapy. Resection may also ultimately prove to be a better option for patients with more severe strictures.

Subendocardial segment length shortening at lateral margins of ischemic myocardium in dogs.

The lateral borders of an infarcted area are sharply delineated in terms of perfusion, but functional impairment extends a limited distance into adjacent nonischemic myocardium. To determine the distribution of functional impairment we arrayed three ultrasonic dimension gauges to measure two subendocardial segment lengths in series. The center crystal, placed at the perfusion boundary (PB) between left anterior descending and circumflex arteries, radiated ultrasound to receiver crystals 7-17 mm to either side of the PB. The locations of the functional measurements relative to the PB were determined with myocardial blood flow (microsphere) "maps" constructed from multiple small tissue samples obtained circumferentially. On the nonischemic side of the PB, segment shortening (dL) increased from 2.00 +/- 0.37 mm during control conditions to 2.20 +/- 0.43 mm (P less than 0.05) after left circumflex coronary occlusion. Similar results were obtained in four conscious chronically instrumented dogs, supporting the conclusion that segment function adjacent to the ischemic margin is well preserved after coronary occlusion. On the ischemic side of the PB, dL decreased from 2.24 +/- 0.54 to 0.42 +/- 0.39 mm (P less than 0.01). By adding the data from the two segments in series, a combined measurement of dL across heterogeneously perfused myocardium was derived that decreased by 38% from control. The level of shortening represented an integral of normal and abnormal motion that was proportional to the mean reduction in blood flow (-44%) in all of the muscle spanned by the crystals. We conclude that subendocardial segment lengths "average" shortening in the muscle they subtend when arrayed across the perfusion boundary.

The functional border zone in conscious dogs.

Studies focusing on the functional border zone have been performed largely with anesthetized, open-chest preparations. Therefore, we instrumented 14 dogs at sterile surgery with sonomicrometers arrayed to measure systolic wall thickening across the perfusion boundary produced by circumflex coronary occlusion. We fitted sigmoid curves to the data to model the distribution of wall thickening impairment as a function of distance from the perfusion boundary, which was delineated with myocardial blood flow (15 micron diameter microspheres) maps. Using this approach, we defined the functional border zone as the distance from the perfusion boundary to 97.5% of the sigmoid curve's nonischemic asymptote. The lateral extent of the functional border zone, measured 10 min and 3 hr after occlusion, was 32 and 28 degrees of circumference, respectively. To evaluate the severity of nonischemic dysfunction, we measured average systolic wall thickening within the functional border zone. It was reduced from 3.69 +/- 1.10 (mean +/- SD) mm to 2.98 +/- 1.07 mm (p less than .01) and 2.74 +/- 1.12 mm (p less than .01) early and late after coronary occlusion. Thus, a narrow functional border zone was evident during circumflex coronary occlusion in conscious dogs. Its lateral extent was limited to approximately 30 degrees (similar to findings in open-chest, anesthetized dogs), severe dysfunction was restricted to the immediate vicinity of the perfusion boundary, and the average severity of nonischemic dysfunction within the functional border zone was mild.

Surgical treatment after the failed antireflux operation.

Eighty-seven adults have undergone reoperation for recurrent gastroesophageal reflux or complications of prior antireflux procedures. Operations performed included the transthoracic Collis-Nissen procedure (59), Collis-Belsey repair (14), Nissen fundoplication (one), repair of acute postoperative paraesophageal hernia (one), division of obstructing crural suture (one), and esophageal resection (23). Among the 73 patients undergoing an additional hiatal hernia repair, there were two postoperative deaths. Follow-up averages 28 months. Subjectively, results have been excellent or good (no or mild reflux symptoms or dysphagia) in 47 (67%); fair in eight (12%), who have moderate dysphagia or reflux symptoms controlled medically; and poor in 15 (21%), 12 of whom have required additional operations. Early postoperative esophageal dilations were required in 25 patients (36%) and regular dilations in seven (10%). Among the 23 patients undergoing esophageal resection, four had a distal esophagectomy and short-segment colon interposition and 19 had a transhiatal esophagectomy without thoracotomy; stomach was used for esophageal replacement in 14 and colon in five. There were no operative deaths. Follow-up averages 17 months. Thirteen patients have had esophageal dilations (nine early and four regularly), and one has clinically significant reflux. Overall, subjective results are good or excellent in 64 (76%). The results of "redo hiatal hernia operation" are far from ideal. Optimal surgical treatment after the failed antireflux operation requires careful analysis of the existing anatomy and experience to decide when esophageal resection is a safer and more reliable approach than another hiatal hernia repair.

Direct measurement of inner and outer wall thickening dynamics with epicardial echocardiography.

Simple geometric models of the left ventricle and indirect experimental measurements suggest that the inner myocardial wall contributes the largest fraction to total wall thickening. We measured transmural differences in regional wall thickening directly, using an epicardial M mode echocardiographic transducer (6 mm diameter, 5 MHz) placed on the anterior free wall of the left ventricle. Wall thickness was partitioned into inner and outer regions by inserting a waxed, 3-0 suture at different depths within the wall. The suture was used as an intramural echo target that was imaged simultaneously with the endocardium to determine inner and outer fractional contribution to total wall thickness. Data were collected in open-chest dogs at rest, during inotropic stimulation with isoproterenol, and during right heart bypass, which was used to vary cardiac output and preload. Results obtained with this method demonstrated that systolic wall thickness was nonuniform at rest and during each intervention. The fractional contributions to total wall thickening of the inner, middle, and outer thirds of the myocardial wall were estimated from the data to be 58%, 25%, and 17%, respectively. The experimental findings corresponded closely to theoretical predictions, supporting the conclusion that a gradient of thickening exists across the myocardial wall, with the inner portion of the wall contributing the largest fraction to total systolic thickening.

The distribution of functional impairment across the lateral border of acutely ischemic myocardium.

To evaluate the degree and lateral extent of dysfunction in nonischemic myocardium adjacent to ischemic muscle, we measured systolic wall thickening with sonomicrometers during circumflex coronary occlusion in 12 anesthetized, open-chest dogs. The locations of the wall thickness measurements relative to the perfusion boundary were determined with myocardial blood flow (microspheres) maps constructed from multiple, small tissue samples. Five minutes after circumflex occlusion, systolic wall thickening in the central ischemic zone decreased from 3.00 +/- 0.61 (mean +/- SD) mm to -0.61 +/- 0.36 mm (P less than 0.01). In nonischemic myocardium greater than 10 mm from the perfusion boundary, systolic wall thickening increased from 2.56 +/- 0.57 to 3.24 +/- 0.72 mm (P less than 0.01). In nonischemic myocardium within 10 mm of the perfusion boundary, systolic wall thickening was slightly but significantly reduced compared with control (2.72 +/- 0.80 to 2.44 +/- 0.79 mm, P less than 0.05), supporting the concept of regional dysfunction in nonischemic myocardium at the lateral borders of an ischemic area. Sigmoid curves were fitted to the data to model changes in wall thickening as a continuous function of distance from the perfusion boundary. This allowed estimation of the extent of dysfunction into nonischemic myocardium which averaged less than 8 mm (approximately 30 degrees of endocardial circumference) at one border. The level of functional impairment in this zone was relatively modest, and systolic wall thickening in the immediate border area was reduced more than 50% from control only in tissue characterized by a blood supply of mixed ischemic and nonischemic origin. We conclude that a functional border zone exists lateral to an acutely ischemic area, but measurement of regional function produces relatively small exaggeration of the size of the acutely ischemic zone if severe reduction in mechanical performance is used to define the extent of the ischemic area.

Massive fatal ventricular septal defect due to nonpenetrating chest trauma in a six-year-old boy: the role of early invasive monitoring in an evolving lesion.

Isolated ventricular septal defects due to nonpenetrating chest trauma are unusual lesions in adults and very rare in children. A review of the literature and of the natural course of traumatic ventricular septal defects is discussed, with emphasis on the evolving course and frequent late appearance of this lesion. Surgical closure is the treatment of choice, most often performed electively following medical treatment. Defects resulting in progressive severe congestive heart failure must be repaired emergently. The key to successful outcome is early suspicion of such a defect in the presence of heart failure following a period of relative cardiovascular stability. Anticipation of evolving myocardial dysfunction in the presence of seemingly minor chest wall injury and nonspecific electrocardiographic changes is emphasized. A case of a fatal traumatic ventricular septal defect is presented. The importance of early invasive monitoring with proper interpretation of hemodynamic data is discussed.