Ischemic Esophageal Necrosis Secondary to Traumatic Aortic Transection

Ischemic Esophageal Necrosis Secondary to Traumatic Aortic Transection

Ann Thorac Surg 2004;78:2175– 8 References 1. McGoldrick JP, Wells FC. Type 1 aortic dissection with right coronary artery occlusion and fistula to r...

191KB Sizes 1 Downloads 22 Views

Ann Thorac Surg 2004;78:2175– 8

References 1. McGoldrick JP, Wells FC. Type 1 aortic dissection with right coronary artery occlusion and fistula to right atrium and right ventricle. Eur J Cardiothorac Surg 1990;4:514 –6. 2. Spier LN, Hall MH, Nelson RL, Parnell VA, Pogo GJ, Tortolani AJ. Aortic dissection: rupture into right ventricle and right pulmonary artery. Ann Thorac Surg 1995;59:1107–9. 3. Kuipers FM, Schatz IJ. Prognosis in dissecting aneurysm of the aorta. Circulation 1963;27:658 –61. 4. Temple TE Jr, Rainey RL, Anabtawi IN. Aortico-atrial shunt due to rupture of a dissecting aneurysm of the ascending aorta. J Thorac Cardiovasc Surg 1966;52:249 –54. 5. Hsu RB, Chien CY, Wang SS, Chu SH. Aorto-right atrial fistula. A rare complication of aortic dissection. Tex Heart Inst J 2000;27:64 –6. © 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc

2175

Ischemic Esophageal Necrosis Secondary to Traumatic Aortic Transection Nam-Hee Park, MD, Jae-Hyun Kim, MD, Dae-Yung Choi, MD, Sae-Young Choi, MD, Chang-Kwon Park, MD, Kwang-Sook Lee, MD, Seong-Wook Han, MD, and Young-Sun Yoo, MD Departments of Thoracic and Cardiovascular Surgery, and Internal Medicine, Keimyung University, Dongsan Medical Center, Daegu, Republic of Korea

Esophageal necrosis with perforation secondary to traumatic aortic transection is extremely rare but usually fatal. A 47-year-old man complained of sudden swallowing difficulty 6 days after blunt trauma. Computed tomography showed a ruptured aorta and the midesophagus shifted to the right side with luminal obliteration because of the ruptured aorta. After primary repair of the partially transected aorta, unexpected mediastinitis because of esophageal perforation was noted. Upper endoscopy showed midesophageal ulceration, necrosis, and perforation. Biopsy samples were consistent with ischemia. The possibility of direct esophageal trauma or intraoperative esophageal injury was ruled out. Esophageal exclusion with thoracoscopic decortication and multiple antibiotics were ineffective, and the patient eventually died. Ischemic esophageal necrosis caused by mechanical compression can occur in a traumatic aortic transection. Dysphagia, when present with radiologic signs, indicates a displaced and compressed esophagus. In spite of aggressive surgical and medical treatment for a perforated esophagus, the prognosis remains poor. (Ann Thorac Surg 2004;78:2175– 8) © 2004 by The Society of Thoracic Surgeons

I

schemic esophageal necrosis after a traumatic aortic rupture is an extremely rare and usually fatal complication [1]. Despite aggressive surgical and medical treatments, all reported cases have died. Here we report esophageal ischemic necrosis related to extrinsic compression by a mediastinal hematoma after a traumatic aortic transection. A 47-year-old male automobile accident victim was admitted to the hospital complaining of right chest and back pains. An initial chest roentgenogram showed only fractures in the right third and fourth ribs and the clavicle. He remained in stable condition for several days

Accepted for publication July 25, 2003. Address reprint requests to Dr Choi, Department of Thoracic and Cardiovascular Surgery, Keimyung University, Dongsan Medical Center, 194 Dongsan-dong, Jung-gu, Daegu, Republic of Korea 700-712; e-mail: [email protected].

0003-4975/04/$30.00 doi:10.1016/j.athoracsur.2003.07.003

FEATURE ARTICLES

tion. However, fistula formation to the cardiac chambers (right atrium, left atrium, and right ventricle) and the pulmonary artery are extremely rare complications of aortic dissection, and are life threatening [1, 2]. They also cause progression of congestive heart failure. Kuipers and Schatz [3] first reported rupture of a chronic dissection into the right atrium in 1963. According to their report, the diagnosis was made at autopsy. Temple and colleagues [4] reported the first successful surgical treatment of aorto-right atrial fistula complicated with aortic dissection. Their case was similar to our case. Initially, the onset of aortic dissection was uncertain. Second, aorto-right atrium fistula was complicated with chronic type A aortic dissection. Third, the symptoms were caused by right side heart failure. However, we have found no previous reports that aorto-right atrial fistula complicated with chronic aortic dissection induced acute liver and renal failure. The aorto-right atrial fistula can be diagnosed by combination of physical examination, cardiac catheterization, TEE, and magnetic resonance imaging. If the patient’s condition is stable, cardiac catheterization is the gold standard to detect aorto-right atrial fistula. However, our case required emergent surgical repair, because his condition was rapidly deteriorating. Therefore, we did not perform cardiac catheterization preoperatively, although we performed TTE first. However, we could not detect aorto-right atrial fistula by TTE. It was difficult to distinguish aortoright atrial fistula from tricuspid regurgitation by TTE [5]. Transesophageal echocardiography is a useful noninvasive method to confirming a diagnosis of aorto-right atrial fistula because it can precisely localize the fistula; Doppler echocardiography can visualize the shunt flow. In addition, TEE can be performed intraoperatively. Our patient was 85 years old. He is our oldest case of aorto-right atrial fistula complicated with aortic dissection. Although progressive hepatorenal failure developed, his postoperative course has been good. In conclusion, aorto-right atrial fistula complicated with aortic dissection is extremely rare and life threatening. When such a diagnosis is made, immediate surgical intervention should be performed. Transesophageal echocardiography is a useful method of confirming the diagnosis.

CASE REPORT PARK ET AL ESOPHAGEAL NECROSIS AFTER AORTIC RUPTURE

2176

CASE REPORT PARK ET AL ESOPHAGEAL NECROSIS AFTER AORTIC RUPTURE

Fig 1. Initial chest computed tomographic scan. The leakage of contrast media compresses the trachea and esophagus. The lumen of the esophagus cannot be seen. FEATURE ARTICLES

with conservative management. On the evening of day 6 of his admission, he complained of dysphagia during medication. Afterwards, he was unable to tolerate liquids. Because he had well tolerated a liquid and solid diet during his admission, we had planned to perform upper endoscopy for the newly onset dysphagia. The next afternoon, however, he complained of a sudden, sharp, substernal chest pain radiating to the back with respiratory distress. He then went into shock and was transferred to the intensive care unit. Cardiac echocardiography showed a left atrium severely compressed by an external mass. A subsequent computed tomographic

Fig 2. The follow-up chest computed tomographic scan. It shows diffuse mediastinitis with abscess and empyema as a result of esophageal perforation.

Ann Thorac Surg 2004;78:2175– 8

Fig 3. Upper endoscopy. Ulceration and necrosis with fistula formation in the midesophagus.

(CT) scan showed contrast leakage in the descending aorta. The trachea and esophagus showed rightward displacement with signs of left atrial compression by the ensuing hematoma. In Figure 1, the hematoma completely occludes the lumen of the midesophagus. After this diagnosis, the patient was referred to Dongsan Medical Center and an emergent operation was performed through a left posterolateral thoracotomy. Because the descending thoracic aorta appeared edematous and discolored by the periadventitial infiltration of the contained rupture, we decided that dissection and repair of the aorta under cardiopulmonary bypass and deep hypothermic circulatory arrest would be safer. After the patient was cooled to a rectal temperature of 18°C, cardiopulmonary bypass was stopped and the aorta was

Fig 4. Microscopic examination shows mostly necrotic tissue composed of fibrinoid material, lymphocytes, neutrophils, smooth muscle, and proliferating small vessels. There is no mucosal gland. (Hematoxylin & eosin, original magnification ⫻40).

opened. On inspection of the aortic lumen, the aortic isthmus was almost completely transected, and only 1 cm of the posterior wall remained intact. The aorta was repaired directly by continuous suture with 3-0 Prolene (Ethicon, Somerville, NJ). The patient had an uneventful postoperative course and was transferred to the general ward 2 days after the operation. However, dysphagia persisted until postoperative day 5, despite surgical decompression. A physical examination was performed to rule out an oropharyngeal injury because of the endotracheal tube placement. Nothing abnormal was found. On postoperative day 6, the drainage tube was removed and the patient started oral feeding on a liquid diet. An intermittent low-grade fever, mild leukocytosis, and an increased C-reactive protein level developed on postoperative day 7. A high fever ensued on postoperative day 10, after which the patient soon became dyspneic and hypotensive and went into shock. A chest roentgenogram showed an air–fluid level in the left hilar area, and the subsequent chest CT scan showed signs of mediastinitis and empyema on the left pleura [Fig 2]. Later on postoperative day 10, a pleural drainage tube was reinserted on the left side, through which 350 mL of dark, bloody and necrotic materials were drained. Because an esophageal perforation had become clinically evident, an esophagography and upper endoscopy were performed on postoperative day 11. The diagnosis was confirmed through the observation of contrast leakage in the midesophagus and visualization of ulceration with necrosis and perforation 22 to 26 cm from the incisor [Fig 3]. The pathologic findings were consistent with ischemia. This diagnosis was further supported by the loss of mucosal glands, smooth muscle necrosis, and the heavy infiltration of inflammatory cells [Fig 4]. Based on the above findings, esophageal exclusion by cervical esophagotomy and division of the gastroesophageal junction was performed on postoperative day 12, along with concomitant jejunostomy feeding and thoracoscopic decortication. The patient’s condition improved after this second operation. On the day 16 after the second operation, however, fresh-colored hemoptysis suddenly developed and profuse, dark, bloody material began passing through the left pleural drainage tube. An aortogram failed to show evidence of any branch bleeding or bleeding from the aortic suture line. Despite spontaneous cessation of bleeding through the pleural tube and improvement in hemoptysis, the patient had to be placed on mechanical ventilation for unremitting hypoxemia. The patient eventually succumbed to mediastinitis on the day 28 after the second operation. An autopsy was not performed.

Comment Esophageal infarction and necrosis can result from various disorders of the aorta. A review of the reported cases suggests two possible modes of injury as being important. One is esophageal ischemia from arterial occlusion

CASE REPORT PARK ET AL ESOPHAGEAL NECROSIS AFTER AORTIC RUPTURE

2177

secondary to aortic dissection [2]. The other is direct extrinsic esophagus compression secondary to a ruptured thoracic aorta or an enlarged aortic aneurysm [1, 3]. A possible explanation for the latter cause is that because the esophagus has a rich intramural arteriovenous network complementing the segmental blood flow, either a localized mediastinal hematoma or an aneurysm enlarged enough to compress the esophagus can obliterate the local venous return and arterial blood supply, leading finally to infarction with perforation [1, 6]. In the present case, the patient had dysphagia and the radiologic signs were compatible with mechanical extrinsic compression of the esophagus by a mediastinal hematoma from the injured aorta. Endoscopy showed mucosal ulcerations with signs of transmural midesophageal necrosis, all of which were findings rarely seen in the absence of other underlying esophageal disease. The microscopic findings were also consistent with esophageal ischemia. These findings strongly suggested that the esophageal necrosis was caused by mechanical extrinsic compression from a mediastinal hematoma after an aortic rupture. Direct esophageal injury during blunt trauma is one possibility in this case. However, esophageal injury that is due to blunt trauma is rare and usually occurs in the cervical esophagus or the gastroesophageal junction. During the initial hospitalization, the absence of any apparent signs of esophageal injury was supported by the patient’s comfortable tolerance of oral intake for up to 6 days after the trauma. Furthermore, the patient had no fever, chest pain, leukocytosis, or mediastinal air shadow, all of which are prominent signs of direct esophageal injury. Therefore, direct injury from blunt trauma can be ruled out as a possible cause of esophageal injury in this case. Intraoperative iatrogenic esophageal trauma was another consideration in the mechanism of injury. However, the risk of injury was minimized by the aorta being repaired under deep hypothermic circulatory arrest with the opened technique, thereby obviating the need for extensive mediastinal dissection and aortic crossclamping. Direct esophageal injury during suturing was also improbable as the posterior one third of the aortic wall was intact. Deep hypothermic circulatory arrest might have contributed to decreasing the esophageal blood flow. However, there are no reports about esophageal infarction after an aortic operation in which deep hypothermic circulatory arrest is used. The fact that the dysphagia started before the operation is the strongest evidence that esophageal injury was not due to operative injury. The exclusion method is a limited treatment option for the management of esophageal injury. A safe and secure primary closure of the fragile and infected esophagus is usually impossible. According to previous reports, this procedure is associated with a very poor prognosis because of death from uncontrollable mediastinitis [1–3]. Well aware of this possibility, we performed concomitant thoracoscopic decortication with drainage and closed irrigation with diluted povidone-iodine solution. The

FEATURE ARTICLES

Ann Thorac Surg 2004;78:2175– 8

2178

CASE REPORT ALTER ET AL VSD AFTER MYOCARDIAL INFARCTION

FEATURE ARTICLES

patient was also prescribed an aggressive antibiotic regimen. However, the uncontrolled mediastinal inflammation had eroded the lung or other mediastinal structures, leading to the final outcome. We are in agreement that our treatment modality did not effectively control the mediastinitis. Therefore, we recommend removal of the perforated esophagus and wide drainage of the mediastinum to avoid the development of an abscess in the perforated esophagus. Because esophageal necrosis secondary to aortic rupture is not anticipated in most cases, the diagnosis is usually delayed. We also initially ignored the symptoms of dysphagia, which did not seem to be related to aortic rupture; therefore, no investigations of dysphagia were performed. We recognized the esophageal perforation after the mediastinitis had spread widely. The only clinical hallmark of the extrinsic compression of the esophagus is newly onset dysphagia or odynophagia [4]. Displacement of the esophageal tube to the right side on plain film [5] or shifting the esophagus with luminal obliteration on a chest CT scan is also helpful in confirming suspected injury. Because upper endoscopy may offer a more accurate means of diagnosis [6], it is recommended that this procedure be performed routinely in patients with one or more suspicious findings of esophagus compression. We believe that patient survival can best be ensured by early recognition and prompt decompression because we do not know what the critical time is for saving the esophagus from irreversible ischemic damage. Notwithstanding, only immediate repair of the ruptured aorta as soon as possible may minimize the ischemic time. In addition, surgical esophageal occlusion before the establishment of widespread mediastinitis may increase the chances of survival in these sick patients. If mediastinitis is widespread, we recommend more aggressive surgical management, including removal of the perforated esophagus with wide drainage of the mediastinal abscess.

References 1. Lacombe M, Coquillaud JP, Andreassian B, Baumann J. Acute traumatic rupture of the thoracic aorta with secondary necrosis of the esophagus. Ann Thorac Surg 1971;11:171–5. 2. Minatoya K, Okita Y, Tagusari O, et al. Transmural necrosis of the esophagus secondary to acute aortic dissection. Ann Thorac Surg 2000;69:1584 –6. 3. Lee KR, Stark E, Shaw FE. Esophageal infarction complicating spontaneous rupture of the thoracic aorta. JAMA 1977;237: 1233–4. 4. Elloway RS, Mezwa DG, Alexander T. Foregut ischemia and odynophagia in patient with a type III aortic dissection. Am J Gastroenterol 1992;87:790 –3. 5. Gerlock AJ Jr, Carlos AM, Coulam CM, Hayes P. Traumatic aortic aneurysm: validity of esophageal tube displacement sign. AJR 1980;135:713–8. 6. Watanabe S, Nagashima R, Shimazaki Y, et al. Esophageal necrosis and bleeding gastric ulcer secondary to ruptured thoracic aortic aneurysm. Gastrointest Endosc 1999;50: 84.7–9. © 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc

Ann Thorac Surg 2004;78:2178 – 80

Long-Term Survival With Acquired Ventricular Septal Defect After Myocardial Infarction Peter Alter, MD, Bernhard Maisch, MD, and Rainer Moosdorf, MD Departments of Internal Medicine–Cardiology and Cardiovascular Surgery, Philipps University of Marburg, Marburg, Germany

Acquired ventricular septal defects (VSD) are rare and devastating complications after myocardial infarction. The long-term prognosis with medical therapy is extremely poor. We report on a patient who developed progressive heart failure within 3 months after myocardial infarction due to an unknown VSD. The left ventricular function was severely impaired. After diagnosing VSD by echocardiography, surgical occlusion was performed. In addition, a biventricular pacemaker was applied using epicardial leads. The patient recovered almost completely 6 weeks postoperative. Beside hemodynamic changes, biventricular pacing is potentially sufficient to improve the postoperative outcome of patients with severe heart failure in these conditions. (Ann Thorac Surg 2004;78:2178 – 80) © 2004 by The Society of Thoracic Surgeons

D

evelopment of ventricular septal defect (VSD) is a rare devastating complication after myocardial infarction with an incidence of 0.2% to 2% [1]. The prognosis remains extremely poor with a mortality of 20% to 50% in patients undergoing surgical therapy and 95% in those treated medically, being highest within the first month after myocardial infarction (overall mortality 74% at 30 days, 78% at 1 year) [2]. Accordingly, patients with long-term survival for several months after medical treatment of myocardial infarction are absolutely rare. Potentially, biventricular pacing is sufficient to improve the postoperative outcome of patients with initially severely reduced left ventricular function. We report a 69-year-old patient who had suffered from angina pectoris. Because of delayed consultation of his physician at the following day, the underlying inferior myocardial infarction has been treated medically without thrombolysis or acute revascularization. Three days later, coronary angiography revealed triple-vessel disease with medial occlusion of the right coronary artery (RCA) as well as high grade stenoses of the left anterior descending (LAD) and circumflex artery (RCX). Angiography Accepted for publication July 25, 2003. Address reprint requests to Dr Alter, Philipps University of Marburg/ Lahn, Department of Internal Medicine–Cardiology, Baldingerstrasse, 35033 Marburg, Germany; e-mail: [email protected].

0003-4975/04/$30.00 doi:10.1016/S0003-4975(03)01449-8