Coselli
J
Si
,
Moreno
P
Li
. Descending and Thoracoabdominal Aneurysm.
In: Cohn LH, Edmunds LH Jr, eds. Cardiac Surgery in the Adult. New York: McGraw-Hill, 2003:1169-1190.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Chapter 48 |
| INTRODUCTION |
|---|
|
|
|---|
| ETIOLOGY AND PATHOGENESIS |
|---|
|
|
|---|
|
Marfan syndrome is a genetic disorder characterized by identifiable connective tissue defects that lead to aneurysm. The aortic wall is weakened by fragmentation of elastic fibers and deposition of extensive amounts of mucopolysaccharides.5 Many patients with Marfan syndrome have an abnormal mutation of the fibrillin gene located on the long arm of the 15th chromosome.6 Abnormal fibrillin in the extracellular matrix decreases connective tissue strength in the aortic wall and produces abnormal elastic properties that predispose the aorta to dilatation from wall tension resulting from left ventricular ejection impulses (DP/DT). Laplace's law causes cycles of progressive dilatation as increasing luminal diameters produce greater wall tension. The usual histologic changes of the aging aorta include cystic medial necrosis, elastin fragmentation, fibrosis with increased collagen, and medial necrosis.7
An adequate preoperative assessment of the diameter of the aorta remains the single most important factor in the decision to repair a thoracic aortic aneurysm. The Ad Hoc Committee on Reporting Standards of the Society for Vascular Surgery and the North American Chapter of the International Society for Cardiovascular Surgery states that the definition of an aneurysm is a permanent localized dilatation of an artery having at least 50% increase in diameter compared to the expected normal diameter of the artery in question.8 This definition can be applied to the thoracic aorta, but it is first necessary to know the normal diameter.
The average diameter of the mid-descending thoracic aorta is 28 mm for men and 26 mm for women; at the level of the celiac axis, 23 mm for men and 20 mm for women; and at the infrarenal aorta, 19.5 mm for men and 15.5 mm for women.9 Normal aortic diameters, however, vary according to age, gender, and body surface area. Aortic enlargement with advancing age is reported in a number of studies.10 Even when corrected for age and body surface area, aortic size is statistically smaller in women than in men. On average, the aorta is 2 to 3 mm greater in diameter for men than for women. Body surface area is a better predictor of aortic size than height or weight and best correlates with aortic diameter in patients less than 50 years of age.11
| NATURAL HISTORY |
|---|
|
|
|---|
Cambria reported 57 patients with nondissecting TAAA preselected for nonoperative management.13 The overall survival rate was 39%, and the repair-free survival rate was only 17% at 5 years. They found an expansion rate of 0.2 cm per year. Dapunt, however, reported the rate of enlargement of thoracic and thoracoabdominal aortic aneurysms in 67 patients followed by computed tomography scanning, and demonstrated an enlargement rate of 0.43 cm per year.14 A significantly higher rate of aneurysm expansion was found in smokers and in patients with larger aortic diameters (more than 5 cm) at diagnosis. No correlation was noted between rate of enlargement and age, sex, or the presence of dissection.
In a series of 53 patients with thoracic aneurysms followed for at least 6 months after diagnosis, Masuda found, by univariate analysis, that three variablesinitial size of the aneurysm, diastolic blood pressure, and presence of renal failurewere statistically correlated to expansion rate of thoracic aortic aneurysms.15 In a population-based study, Bickerstaff reported the outcome of thoracic aortic aneurysms in 72 patients.16 Rupture occurred in 53 patients (74%) and 50 died. The median interval between diagnosis and rupture in 16 patients with known aneurysms was 2 years. Ninety-five percent of aortic dissections ruptured and 51% of nondissecting aneurysms ruptured. The actuarial 5-year survival for all 72 patients was 13%; for patients with aortic dissection it was 7%, and for those without dissection, 19.2%.
Pressler and McNamara compiled data from 176 patients with the diagnosis of thoracic aortic aneurysm.17 In their study, 90 patients (51%) had arteriosclerotic fusiform aneurysms and 86 (49%) had dissecting aneurysms. Eighty-nine percent of the arteriosclerotic aneurysms and 41% of the dissecting aneurysms were located in the descending thoracic aorta. Seventy-eight percent of patients with dissecting aneurysms had symptoms of pain in the back or chest, whereas only 42% of patients with atherosclerotic aneurysms had back or chest pain. Rupture caused 37 of 48 deaths (77%) in 59 patients with dissecting aneurysms, and 25 of 57 deaths (44%) in 76 patients with atherosclerotic aneurysms. Concomitant cardiovascular disease was the second leading cause of death in patients with atherosclerotic aneurysms not treated surgically (19 deaths out of 86 patients; 22%), but only one of 59 patients with dissecting aneurysms not treated surgically died of heart disease.
Aortic size at the time of diagnosis is related to the development of complications including rupture. Pressler and McNamara reported that 8 of 9 ruptured descending thoracic aortic aneurysms were larger than 10 cm.17 In the report by Dapunt on descending thoracic and thoracoabdominal aortic aneurysms, mean aortic diameter at the time of rupture was 6.1 cm.14 Crawford found a mean size of 8 cm at the time of rupture in a series of 117 patients with descending thoracic and thoracoabdominal aortic aneurysms. Since rupture was observed in some 10% of aneurysms smaller than 6 cm in diameter, the authors recommended elective operation when a 5-cm diameter threshold was exceeded.18 Factors associated with increased morbidity and mortality from rupture and related complications include aneurysm size, presence of dissection, previous rate of expansion, and hypertension (Table 48-2).
|
Increasing age and preoperative renal insufficiency have remained major risk factors for early mortality throughout the history of TAAA repair. Both were among the predictive variables determined by Svensson et al's multivariable analysis of Crawford's complete experience with TAAA surgery in 1509 patients treated between 1960 and 1991.20 The recent report by Acher et al confirms that, along with acute presentation, age and elevated creatinine levels remain important predictors of early death.21
| CLINICAL PRESENTATION |
|---|
|
|
|---|
Compression of the trachea or bronchus can produce stridor, wheezing, or cough. Pneumonitis distal to an area of bronchial obstruction develops if secretions cannot be cleared. Hemoptysis occurs when an aneurysm erodes directly into the pulmonary parenchyma or bronchus. Compression of the esophagus may produce dysphasia, whereas erosion into the esophagus causes hematemesis.22 Similarly, erosion into the duodenum causes either partial obstruction or intermittent/massive gastrointestinal bleeding. Compression of the liver or porta hepatis is uncommon, but when this occurs, jaundice results. Hoarseness is owing to traction on the vagus nerve as the distal aortic arch expands, to produce recurrent laryngeal nerve paralysis. Thoracic or lumbar vertebral body erosion causes back pain, spinal instability, and neurologic deficits from spinal cord compression. Mycotic aneurysms have a peculiar propensity to destroy vertebral bodies. Additionally, neurologic symptoms, including paraplegia and/or paraparesis, may occur with thrombosis of intercostal and spinal arteries. This is most frequently seen with acute aortic dissection, which may occur primarily, or become superimposed on medial degenerative fusiform aneurysmal disease. Erosive fistula formation into the inferior vena cava or iliac vein will present with an abdominal bruit, widened pulse pressure, edema, and heart failure. Thoracic aortic aneurysms, similar to aneurysms in other locations, may produce distal emboli of clot or atheromatous debris that gradually obliterates and thromboses visceral, renal, or lower extremity branches. Secondary infection of atheromatous debris and clot within an aneurysm may produce generalized sepsis. Nine percent of patients with TAAA present with frank rupture at the time of diagnosis.4
| DIAGNOSTIC EVALUATION |
|---|
|
|
|---|
|
Ultrasonography, although useful in evaluating infrarenal abdominal aortic aneurysms, is not useful for imaging the thoracic or suprarenal aorta primarily because of overlying lung tissue.24 The advantages of ultrasonography are wide availability, low cost, portability, noninvasiveness, lack of ionizing radiation, and rapid examination. When the definitive neck of an infrarenal abdominal aortic aneurysm cannot be demonstrated at the level of the renal arteries, thoracoabdominal aortic involvement should be suspected.
Transesophageal Echocardiography
Transesophageal echocardiography provides access to the proximal aorta, and complements transabdominal ultrasonography.25,26 The technique requires considerable technical skill both in obtaining adequate images and in interpretation. The technique is excellent for determining the presence of dissection but has limitations in evaluating the region of the transverse aortic arch and upper abdominal aorta.
Computed tomography scanning is widely available and provides access to the entire thoracic and abdominal aorta. In addition to diagnosis, information regarding location and extent is provided.27 Major branch vessels including the celiac, superior mesenteric, renal, and iliac arteries, left subclavian, and virtually all adjacent organs are imaged. Although not widely available, computer programs can construct sagittal, coronal, and oblique images as well as three-dimensional reconstructions.28,29 Computed tomography scanning, which is contrast-enhanced, provides information regarding the aortic lumen, mural thrombus, presence of aortic dissection, intramural hematoma, mediastinal or retroperitoneal hematoma, aortic rupture, and periaortic fibrosis associated with inflammatory aneurysms (Fig. 48-2).30 Although angiography remains the "gold standard" for evaluating aortic occlusive disease, improvements in computed tomography (CT) and magnetic resonance imaging (MRI) are leading to strategies that provide excellent images without the morbidity or cost of angiography.31 Because of improvements in noninvasive imaging modalities and a stroke risk of 0.6% to 1.2% with angiography, the role of diagnostic angiography for aortic arch vessels is becoming limited.3234
|
Magnetic Resonance Angiography
An important advantage of magnetic resonance angiography (MRA) over computed tomography angiography (CTA) is that it uses nontoxic gadolinium instead of nephrotoxic contrast. Additionally, the patient avoids exposure to ionizing radiation. MRI employs radiofrequency energy and a strong magnetic field to produce images. MRA provides the same volume of information as does CTA with regards to image processing, but further provides information on relative quantity of blood flow and an appearance similar to conventional angiography. Additionally, the technique can provide a three-dimensional anatomical analysis. MRA imaging of the aorta can elucidate information on wall composition, wall thickness, and intraluminal thrombus, whereas conventional aortography only depicts the lumen. A current limitation of MRA is the susceptibility to artifacts created by ferromagnetic materials. Although expensive, the technology is widely available and has the capability of accessing the entire aorta. MRA images can more clearly distinguish arteries and veins from viscera and other surrounding tissue.36
Classical aortography remains the mainstay for preoperative evaluation of patients with thoracoabdominal aortic aneurysms.24 It has the ability to define the extent of aneurysm, branch vessel involvement, and branch vessel stenotic lesions. Risks of aortography include renal toxicity from the large volumes of contrast material required to adequately fill large aneurysms. There is the additional risk of embolization from laminated thrombus secondary to manipulation of intraluminal catheters. Anterior, posterior, oblique, and lateral views are obtained simultaneously to obtain satisfactory information regarding branch vessels. Patients with suspected renal and/or visceral ischemia, aorto-iliac occlusive disease, horseshoe kidney, or peripheral aneurysms should be considered for aortography prior to TAAA repair.
Aortography is performed in a well-hydrated patient who is also receiving intravenous fluids. Routinely, 1000 mL of 5% dextrose and Ringer's lactate solution with 25 g of mannitol are given intravenously immediately prior to the procedure and are continued at 100 mL per hour following study. If at all possible, operation is delayed for 24 hours or longer to determine the effects of angiography on renal function and to permit diuresis of the contrast agent. If renal insufficiency occurs or is worsened, the surgical procedure is postponed until renal function returns to normal or is satisfactorily stabilized.
| PREOPERATIVE ASSESSMENT AND PREPARATION |
|---|
|
|
|---|
A history of coronary artery occlusive disease is present in 30% of patients with thoracoabdominal aortic aneurysms. Additionally, cardiac disease is responsible for 49% of early deaths and 34% of late deaths.3,4 Transthoracic echocardiography is a satisfactory noninvasive screening method that evaluates both valvular and biventricular function. Dipyridamole-thallium myocardial scanning identifies regions of myocardium that are reversibly ischemic and is more practical than exercise testing in this generally elderly population that commonly is limited by concurrent lower extremity peripheral vascular disease. We routinely employ preoperative screening of all patients for coronary artery disease with cine arteriography; however, in patients with a significant history of angina or an ejection fraction of 30% or less, cardiac catheterization and coronary arteriography are performed with aortography. Patients who have asymptomatic TAAA and severe coronary artery occlusive disease (left main, triple vessel, and proximal left anterior descending) undergo myocardial revascularization prior to aneurysm replacement. In appropriate patients, percutaneous transluminal angioplasty is carried out prior to operation.
Renal function is assessed preoperatively by serum electrolytes, blood urea nitrogen (BUN), and creatinine measurements. Renal size may be determined from a CT scan, by ultrasound, or from the nephrogram obtained during aortography. Renal artery patency is confirmed by arteriography.38 Patients are not rejected as surgical candidates based on renal function. Patients with preoperative renal failure and an established hemodialysis program do not have significantly greater morbidity than patients with normal renal function. Patients with severely impaired renal function, but who are not on chronic hemodialysis, frequently require transient temporary hemodialysis early after operation. Additionally, patients with poor renal function secondary to severe proximal renal occlusive disease are revascularized at operation by either renal arterial endarterectomy or bypass grafting with the expectation that renal function will stabilize or improve.
All patients undergo pulmonary function screening with arterial blood gases and spirometry.3941 Patients with an FEV1 greater than 1.0 and a PCO2 less than 45 are surgical candidates. In suitable patients, borderline pulmonary function frequently is improved by stopping smoking, progressively treating bronchitis, losing weight, and following a general exercise program for a period of 1 to 3 months before operation. However, operation is not withheld in patients with symptomatic aortic aneurysms and poor pulmonary function. In such patients, preservation of the left recurrent laryngeal nerve, phrenic nerve, and diaphragmatic function is particularly important.
| OPERATIVE TREATMENT |
|---|
|
|
|---|
Successful conduct of operation requires close coordination between surgeon and anesthesiologist. Advances in anesthetic techniques, monitoring, and perfusion technology have contributed to improved results in the treatment of TAAA. As a result of advanced age and concomitant prevalence of associated coronary artery occlusive disease, anesthesia is induced with narcotic agents (fentanyl) to minimize the risk of myocardial depression. A large-bore central venous line (a three-lumen 12-gauge catheter) and a Swan-Ganz pulmonary artery catheter are placed for access and monitoring. A right radial and, frequently, bilateral radial intra-arterial catheters are placed for monitoring and blood withdrawal. Muscle relaxation is achieved and maintained with pancuronium bromide. Either a double-lumen endobronchial tube is placed for selective ventilation of the right lung and deflation of the left lung, or alternatively a single-lumen endotracheal tube with an intrabronchial blocker is utilized. Deflation of the left lung reduces retraction trauma to the lung, improves exposure, and alleviates the risk of cardiac compression. The patient is turned to a right lateral decubitus position with the shoulders placed at 60° to 80° and the hips flexed to 30° to 40° from horizontal. The position is stabilized using a beanbag. Arterial blood gases, electrolytes, and serum glucose are monitored frequently (3060 minutes). Intraoperatively the electrocardiogram, arterial and venous blood pressures, and temperature are monitored continuously. In patients with a significant history of cardiac disease and/or known impaired cardiac function, a transesophageal echocardiography probe is inserted following induction of anesthesia.
Shortly after the induction of anesthesia, 25 to 50 g of mannitol are given intravenously to promote a vigorous diuresis. Intravenous crystalloid solutions are begun prior to operation. The first liter consists of lactated Ringer's solution with 5% dextrose, and the remainder, Ringer's solution without dextrose in sufficient volumes to maintain the central venous pressure between 7 and 10 mm H2O and the pulmonary capillary wedge pressure at normal or preanesthetic levels. Proximal blood pressure, cardiac hemodynamics, and peripheral vascular resistance are maintained at optimal levels by administration of sodium nitroprusside and/or nitroglycerin, and replacement of fluid and blood losses. Nitroprusside is specifically discontinued several minutes before the release of the distal aortic cross-clamp. Sodium bicarbonate solution is administered routinely at a rate of 2 to 3 mEq/kg/h by continuous infusion during aortic cross-clamping to prevent acidosis.
Throughout the procedure, hemoglobin and coagulation parameters are monitored carefully and are adjusted primarily by replacing appropriate blood components. In general, we administer fresh frozen plasma continuously throughout the operation, and at least one pheresis unit of platelets at the time of aortic declamping. This minimizes problems related to coagulopathy produced by dilution of coagulation proteins. A cell-saving device is used throughout the procedure to salvage all shed blood from the operative field. If necessary, during a period of substantial blood loss, the device allows direct reinfusion of unwashed blood from the reservoir. The authors' preference is to use citrate rather than heparin in the autotranfusion device, and so intermittent monitoring of the serum calcium level is important. Mechanized rapid transfusion devices using large-bore central venous catheters for access are particularly valuable in restoring blood volume immediately prior to declamping.
Heparin is administered intravenously (1 mg/kg) before placing the aortic cross-clamp or initiating left heart bypass. Potential benefits of heparinization include preservation of the microcirculation and prevention of embolization; the authors have not encountered increased bleeding or other morbidity related to the heparin. Following this low heparin dose, the activated clotting time generally ranges from 220 to 270 seconds. By avoiding the initiation of the clotting cascade, the use of heparin may have a favorable influence on reducing the incidence of diminished intravascular coagulation.
Classification of Thoracoabdominal Aneurysms
Thoracoabdominal aneurysms can involve the entire thoracoabdominal aorta from the origin of the left subclavian artery to the aortic bifurcation or can involve only one or more segments. The Crawford classification has advanced the surgical treatment of TAAA because it has permitted a standardized reporting of aneurysm extent, allowing for an appropriate stratification of risks, specific treatment modalities based on the extent of the aneurysm, and a type-specific determination of neurologic deficit as well as morbidity and mortality associated with thoracoabdominal aneurysm repair (Fig. 48-3). Extent I thoracoabdominal aortic aneurysms involve most of the descending thoracic aorta from the left subclavian artery down to vessels in the abdomen. Usually the renal arteries are not involved in extent I aneurysms. Extent II aneurysms begin at the left subclavian artery and reach the infrarenal abdominal aorta even as far as the inguinal area. Extent III aneurysms involve the distal half or less of the descending thoracic aorta and substantial segments of the abdominal aorta. Extent IV aneurysms are those that involve the upper abdominal aorta and all or none of the infrarenal aorta.
|
A fundamental principle is the importance of adequate exposure. The thoracoabdominal incision varies in length and level, depending on the anticipated extent of aortic replacement (Fig. 48-4A). When the aneurysm extends into the superior aspect of the thorax (Crawford extents I and II), the upper portion of the thoracoabdominal incision is through the 6th intercostal space or the bed of the resected 6th rib. In recent years, we have routinely removed a rib. When the interspace is used, the upper rib may be divided at the neck for additional proximal exposure. With lower aneurysms (Crawford extents III and IV), an incision through the 7th, 8th, or 9th interspace is employed according to the desired level of exposure. A straight transverse incision through the 10th or 11th interspace is used in patients with aneurysms between the diaphragm and aortic bifurcation (Crawford extent IV). In all others, a gentle curve to reduce the risk of tissue necrosis at the apex of the lower portion of the musculoskeletal tissue flap is made as the incision crosses the costal margin. In patients with proximal aneurysms, the posterior portion of the incision is located between the scapula and the spinal processes. The distal extent of the incision is carried down to the level of the umbilicus.
|
After entering the chest, the left lung is deflated. Fixed metal retractors attached to the operating table provide consistent static exposure. The diaphragm is divided in a circular fashion to protect the phrenic nerve and to preserve as much diaphragm as possible. Only a 1- to 1.5-cm rim of diaphragmatic tissue is left laterally for closure at the completion of operation.
The abdominal aortic segment is exposed using a transperitoneal approach; the retroperitoneum is entered lateral to the left colon.42 A dissection plane is developed in the retroperitoneum anterior to the psoas muscle and posterior to the left kidney. This is extended directly to the left posterolateral aspect of the abdominal aorta. The left colon, spleen, left kidney, and tail of the pancreas are retracted anteriorly and to the right. An open abdominal approach permits direct inspection of the bowel, abdominal viscera, and visceral blood supply following completion of the aortic reconstruction. An entirely retroperitoneal approach is used in patients with a so-called hostile abdomen, defined by multiple prior abdominal operations, or a history of extensive adhesions and/or peritonitis.
The crus of the diaphragm is divided and the left renal, superior mesenteric, and celiac arteries are identified but not circumferentially dissected or encircled with tapes. Commonly, a large lumbar branch of the left renal vein courses posteriorly around the aorta. This may be ligated and divided as needed. If a retroaortic left renal vein is encountered, the vessel is divided between vascular clamps if the aortic repair extends below the vein. Direct reanastomosis or interposition grafting of this retroaortic renal vein is necessary if the left kidney appears congested with distended testicular, ovarian, and adrenal collaterals.
Patients with extensive thoracoabdominal aortic aneurysms (Crawford extents I and II), and particularly those with dissection, are at greatest risk for development of postoperative paraplegia and paraparesis. In such patients, distal aortic perfusion during the proximal aortic portion of the repair is achieved by using temporary bypass from the left atrium to either the femoral artery (most commonly the left) or distal descending thoracic aorta with a closed-circuit in-line centrifugal pump (Biomedicus, Medtronic, Inc., Eden-Prairie, MN) (Fig. 48-4B). If the pericardium was previously entered for coronary artery bypass grafting or valve replacement (Fig. 48-4C), cannulation of the superior or inferior pulmonary vein is equally effective. Cannulation of the distal descending thoracic aorta (usually at the level of the diaphragm) was initially used solely as an alternative to femoral artery cannulation in patients with femoral or iliac artery occlusive disease. Because of the lack of complications using this technique and the elimination of femoral artery exposure and repair, distal aortic cannulation has become the preferred approach. Careful examination of CT or MRI scans assists selection of an appropriate site for direct aortic cannulation and avoidance of intraluminal thrombus with potential for distal embolization. Bypass flows are adjusted to maintain distal arterial pressures of 70 mm Hg, while maintaining normal proximal arterial and venous filling pressures. Flows between 1500 and 2500 mL/min are generally required. Left heart bypass (LHB) flows are targeted toward two thirds of the baseline cardiac output, which is routinely measured shortly after induction. LHB facilitates rapid adjustments in proximal arterial pressure and cardiac preload, thereby reducing the need for pharmacologic intervention. The patient's temperature is allowed to drift down to a rectal temperature of 32°C to 33°C.
When the aneurysm encroaches on the left subclavian artery, the distal aortic arch is mobilized gently by dividing the remnant of the ductus arteriosus. The vagus and recurrent laryngeal nerves are identified. The vagus nerve may be divided below the recurrent nerve to provide additional mobility, consequently protecting it from injury. Preservation of the recurrent laryngeal nerve is particularly important in patients with chronic obstructive pulmonary disease and reduced pulmonary function. Vocal cord paralysis should be suspected in patients with postoperative hoarseness and confirmed by direct examination. Effective treatment can be provided by direct cord medialization (type 1 thyroplasty), or, in higher-risk patients, by polytetrafluoroethylene injection.43 Careful circumferential dissection of the distal transverse aortic arch separates it from the pulmonary artery and esophagus. If cross-clamping proximal to the left subclavian artery is anticipated, the left subclavian artery is separately and circumferentially mobilized. In patients with a prior left internal mammary arterial bypass graft, either a left common carotid to subclavian bypass or a left subclavian to carotid transfer is necessary to avoid cardiac ischemia when the cross-clamp is applied proximal to the left subclavian artery.
The distal clamp is placed between T4 and T7. Distal aortic perfusion provides circulation to the viscera, kidneys, lower extremities, and lower intercostal and lumbar arteries. The aorta is transected 1 cm distal to the proximal clamp and separated from the esophagus to permit full-thickness sutures through the aortic wall without injuring the esophagus (Fig. 48-4C). A gelatin-impregnated woven Dacron graft (Sulzer Vascutek, Scotland) is selected; 22-mm to 24-mm grafts are used in most patients. The anastomosis and all remaining anastomoses are usually made with a running 3-0 polypropylene suture (Fig. 48-4D). Teflon felt strips are generally not used. In patients with particularly fragile aortic tissues such as those found in Marfan syndrome, 4-0 polypropylene sutures are used. As the aorta is replaced from proximal to distal, the distal aortic clamp is moved sequentially to lower positions along the aorta to maintain distal perfusion and restore proximal blood flow (Fig. 48-4EH). Sequential distal clamping is often not feasible due to a variety of factors related to the severity of the aortic disease, including aneurysm size and tortuosity, mural calcification, and intraluminal thrombus.
Commonly, atrio-distal bypass is discontinued following completion of the proximal anastomosis. The entire aneurysm then is opened longitudinally, passing posterior to the left renal artery and continuing to the distal extent of the aneurysm. A distal clamp is not used, allowing for an "open" anastomosis. With a chronic dissection, the partition between the true and false lumens is completely removed. Aorto-visceral bypass then is restarted using a Y line off the arterial perfusion line and separate balloon perfusion catheters placed within the origin of the celiac, superior mesenteric, and renal arteries. This provides oxygenated blood to the abdominal viscera and kidneys (Fig. 48-5). With this technique, the total renal and visceral ischemic time can be reduced to just a few minutes during even the most complex aortic reconstructions. The potential benefits of reducing hepatic and bowel ischemia include decreased risks for postoperative coagulopathy and bacterial translocation, respectively.
|
In patients with lower aortic aneurysms (i.e., Crawford extents III and IV) atrio distal aortic bypass may be modified to provide only atrio visceral and/or renal bypass (Fig. 48-6). This technique avoids distal aortic or femoral cannulation, but reduces cardiac preload, protects the renal parenchyma, reduces post-clamp acidosis, and may reduce the risk of postoperative bacterial translocation by reducing bowel ischemia.
|
|
|
Following completion of aortic repair, protamine sulfate is administered to reverse heparin. It is imperative that adequate hemostasis is achieved and secured at all suture lines. The renal, visceral, and peripheral circulation are assessed. The aneurysmal wall is then loosely wrapped around the aortic graft. Two posteriorly located thoracic drainage tubes and a closed-suction retroperitoneal drain are placed prior to closure. The diaphragm is closed with running nonabsorbable suture; disruption postoperatively is exceedingly rare. To stimulate and maintain renal function, a low-dose dopamine drip of 2 to 3 mg/kg/min is initiated and continued for 24 to 48 hours. Patients are generally weaned from the respirator overnight and extubated the following morning. All drains are removed and antibiotics discontinued at 36 to 48 hours postoperatively. Ambulation is started on the second postoperative day.
The author routinely uses CSF (cerebrospinal fluid) drainage in patients with Crawford extent I or II. In patients in whom this is considered necessary, an 18-gauge intrathecal catheter is placed through the second or third lumbar space. The catheter permits aspiration of cerebrospinal fluid and pressure monitoring throughout the operation, and is continued for 2 to 3 days postoperatively. CSF is allowed to passively drain from the catheter. CSF is aspirated as needed during the period of aortic occlusion to keep the CSF pressure at or below 10 mm Hg using a closed collection system.
Endovascular Repair of Thoracic Aortic Aneurysms
Parodi et al reported the first clinical use of stent grafts for repair of abdominal aortic aneurysm in 1991.46 Dake et al reported the first endovascular thoracic aortic repair in 1994.47 Subsequent reports have supported this new less invasive therapy for traumatic, mycotic, and ruptured aneurysms of descending thoracic aorta.4850 Theoretically, endovascular repair of thoracic aortic aneurysm with stent grafts is a feasible alternative to standard surgical procedures in the treatment of selected patients with compromised cardiac, pulmonary, or renal status, persons who have undergone previous complex thoracic aortic procedures, and the very elderly. Conceptually, there is the potential to offer potentially reduced operative risk, hospital stay, and procedural cost.
Appropriate anatomy for endovascular repair includes the following: (1) a normal arterial segment proximal to the aneurysm and distal to the left common carotid artery of at least 2 cm in length and less than 38 mm in diameter; (2) a normal arterial segment distal to the aneurysm and proximal to the celiac axis of at least 2 cm in length and less than 38 mm in diameter; and (3) iliac arteries greater than 8 mm in diameter.51
The mortality and morbidity of endovascular repair of the thoracic aortic aneurysm are very difficult to determine. Most of the reported series are small and selection of patients may play a major role in determining outcome. The largest series is from the Stanford group, which included 103 patients. They reported a 30-day mortality rate of 9%, paraplegia/paraparesis (3%), myocardial infarction (2%), respiratory insufficiency (12%), and a high incidence of stroke (7%).52 Balm et al, in a selected group of 144 patients with descending aortic aneurysms treated with endovascular stents, identified an early incidence of endoleak of 23% as documented by angiography or CT scanning.53 The initial experience with these devices suggest that transluminally placed endovascular stent grafts are an attractive alternative for conventional surgery. However, there are not enough data available to determine the long-term effectiveness of this treatment; that will be established by well-controlled, large-scale studies that compare the open operative repair to this new technique.
| PARAPLEGIA AND SPINAL CORD PROTECTION STRATEGIES |
|---|
|
|
|---|
Irreversible paraplegia is one of the most devastating complications after TAAA repair. The incidence of paraplegia or paraparesis, as reported in the literature, following thoracoabdominal aortic aneurysms varies substantially and ranges from 4% to 32%.4 Svensson et al's landmark report of Crawford's experience documented on overall 16% incidence of paraplegia or paraparesis; complete paralysis occurred in more than half of patients with deficits.3 In the author's report of 1108 patients who underwent elective repair, the combined incidence of paraplegia/paraparesis was 3.6% (40 of 1099 patients, excluding 7 patients with preoperative paraplegia and 2 patients who died during operation).54 Generally, in large series the incidence of paraplegia and paraparesis is equally divided.2,3 Up to 30% of patients who develop postoperative neurologic deficits initially awake with lower extremity function but develop deficits subsequently, i.e., "delayed paraplegia."3 Operative factors that contribute to spinal cord injury include the duration and degree of ischemia, reperfusion injury, and loss of critical intercostals and lumbar arteries due to ligation, embolization, or thrombosis.55 The risk of spinal cord injury averages, based on the Crawford classification, 13% for extent I, 28% to 31% for extent II, 7% for extent III, and 4% for extent IV.4 Although in the past aortic dissection was identified as a risk factor, in more recent experience, dissection is no longer a risk factor for the development of postoperative paraplegia or paraparesis.56 This is primarily a consequence of aggressive reattachment of intercostal arteries in patients with aortic dissection. This effort to reattach intercostal arteries has also likely reduced the risk of delayed paraplegia.
Vascular Anatomy of the Spinal Cord
The anatomy of the blood supply to the spinal cord is relevant to prevention of spinal cord ischemia and its sequelae. The major arterial circulation to the spinal cord is the anterior longitudinal spinal artery and the paired posterior longitudinal spinal arteries.57 These vessels originate from intracranial vertebral arteries, or branches thereof, and course along the spinal cord for its entire length. The segmental spinal arteries supplying the thoracic and lumbar regions of the cord originate from the posterior branches of the intercostal and lumbar arteries, respectively. The anatomy is highly variable from one individual to another. The segmental spinal arteries give rise to the large anterior and smaller posterior radicular arteries. Each then directly supplies the anterior and posterior longitudinal spinal arteries. Not all anterior and posterior radicular arteries, however, reach the cord. It is this fact and the fact that the anterior spinal artery frequently is attenuated, or entirely discontinuous, that makes the spinal cord highly vulnerable to ischemia. The artery of Adamkiewicz is the largest of the radicular medullary arteries.58 It has a variable origin, arising between T5 and T8 in 12% to 15% of the cases, between T9 and T12 in 60%, at L1 in 14%, at L2 in 10%, at L3 in 1.4%, and between L4 and L5 in 0.2%. The arteria radicularis magna anterior is a decisive factor influencing spinal cord damage during aortic occlusion.59 When the vessel reaches the anterior spinal artery, generally it bifurcates into a smaller ascending branch and a larger descending branch. Intimal atherosclerosis, particularly in medial degenerative fusiform aneurysms, obliterates many intercostal and lumbar arteries, and complicates matters anatomically.
Pathophysiology of Aortic Cross-Clamping
An understanding of the pathophysiological mechanisms involved in aortic cross-clamping and unclamping is imperative in selecting effective measures to prevent and treat the consequences. Most clinical studies indicate that cardiac output decreases with thoracic aortic cross-clamping, whereas most animal studies show no significant change. The normal heart can withstand large increases in afterload without significant ventricular distention or dysfunction. Although impaired myocardial contractility and reduced coronary reserve are rare in animal experiments, such disorders are frequent in the elderly population undergoing aortic reconstruction.60 Clamping the aorta increases impedance to aortic flow, increases systemic vascular resistance and afterload, and redistributes blood volume because venous vasculature distal to the aortic clamp collapses and constricts. This effectively increases preload. The increases in afterload and preload demand an increase in myocardial contractility, which causes an autoregulatory increase in coronary blood flow. Impaired subendocardial perfusion caused by high intramyocardial pressure, with resultant acute deterioration in left ventricular function and/or myocardial ischemia, may be the cause of wall motion abnormalities and changes in ejection fraction. If coronary blood cannot increase, cardiac decompensation follows.
Pathophysiology of Spinal Cord Ischemia
Injurious effects to the spinal cord, kidneys, lungs, and abdominal viscera are caused primarily by ischemia and reperfusion of organs distal to the aortic clamp and to a release of mediators from ischemic and reperfused organs. The most challenging and troublesome complication following TAAA replacement remains spinal cord injury and the development of paraplegia or paraparesis.
Spinal cord ischemic injury is the result of permanent or temporary interruption of spinal cord blood supply during aortic cross-clamping and permanent disruption of delicate and variable arteries to the spinal cord. Several pathogenetic mechanisms are related to neuronal cell death after transient spinal cord ischemia: excitotoxicity, intracellular calcium overload, nitric oxide, eicosanoids, apoptosis, inflammation, and reactive oxygen species. These mechanisms should be regarded as pathways that act in both parallel and sequential manners.61 The duration of cross-clamping influences the magnitude of spinal cord ischemia and reperfusion. Studies suggest that cross-clamping for a period of less than 30 minutes is frequently safe.6266 Thoracic aortic occlusion results in increased intracerebral blood flow, which contributes to the increased CSF pressure and decreased spinal cord perfusion pressure.67 By reducing CSF pressure, therefore, CSF drainage theoretically improves spinal cord perfusion during periods of thoracic aortic clamping.
An alternative explanation proposed by Piano and Gewertz postulates that increased CSF pressure during aortic clamping is related to volume changes in the venous capacitance beds located in the dural space.68 Based on this model, the benefit of CSF drainage may be related to enhanced patency of these intramural veins. Other authors have suggested that the protective effect of CSF drainage may be attributable to the removal of negative neurotrophic factors that accumulate in the CSF during the ischemic period. Brock et al, for example, observed a strong positive relationship between elevations of CSF excitatory amino acid levels (i.e., glutamate, aspartate, and glycine) during aortic cross-clamping and reperfusion, and subsequent development of clinical signs of spinal cord injury.69 Our recent prospective randomized trial focused solely on the impact of CSF drainage in preventing neurologic deficits after TAAA repair. The control and treatment groups were extremely well matched and a consistent surgical strategy was used throughout the study. The trial clearly showed that CSF drainage prevents paraplegia after the repair of extent I and II TAAA.70
Sodium nitroprusside during thoracic aortic cross-clamping reduces spinal cord perfusion pressure and increases the incidence of neurologic deficits.71 The decrease in cord perfusion pressure is owing to a decrease in the distal aortic pressure beyond the clamp and an increase in CSF pressure. The increase in CSF pressure occurs from cerebrovasodilatation.72 Drugs to reduce proximal aortic pressure ideally should possess minimal cerebrovasodilating properties.
The neuroprotective effect of hypothermia is presumed to be secondary to decreased tissue metabolism and a generalized reduction in energy-requiring processes in the cell. However, the mechanisms may be more complex and involve membrane stabilization and reduced release of excitatory neurotransmitters.73 The author uses mild passive hypothermia (31°C-33°C) in all cases. Frank et al report a technique using partial bypass and moderate hypothermia for organ protection during the clamp-induced ischemic period.74 The advantages of moderate over deep hypothermia include a stable intrinsic cardiac rhythm that eliminates the need for full cardiopulmonary bypass. They report a series of 18 patients undergoing thoracic and thoracoabdominal aortic aneurysm resection and replacement with moderate (30°C) hypothermia and partial bypass (aorto-femoral or atrio-femoral). No patient developed paraplegia or significant renal failure. There were two deaths (11%). The advantages of moderate over deep hypothermia include a stable intrinsic cardiac rhythm that eliminates the need for full cardiopulmonary bypass. Most authors specifically avoid the technique of profound hypothermia and circulatory arrest for TAAA repair, principally because of the threat of coagulopathy, pulmonary dysfunction, and massive fluid shift.
Crawford et al reported the clinical use of cardiopulmonary bypass using hypothermic circulatory arrest in 25 patients treated for thoracic aortic aneurysms through a posterolateral approach.75 There were 21 early survivors and cerebral protection was entirely satisfactory. The technique was not entirely effective in eliminating paraplegia; 2 (11%) of 18 patients at risk for ischemic spinal cord injury developed neurologic deficits. This may be explained by satisfactory cord protection during the period of ischemia, but spinal cord injury from sacrifice of critical intercostal arteries.
Kouchoukos et al have recently reported on the use of hypothermic cardiopulmonary bypass with circulatory arrest as an adjunct for operations on the distal aortic arch, descending thoracic aorta, and thoracoabdominal aorta.76 They evaluated 161 patients. Their 30-day mortality rate was 6.2%, and 90-day mortality rate was 11.8%. Paraplegia occurred in 4 and paraparesis in 1 of 156 operative survivors. Renal dialysis was required in 4 (2.5%). They identified hypothermic cardiopulmonary bypass as providing safe and substantial protection against paralysis and renal, cardiac, and visceral organ system failure.
There are two variations of regional spinal cord hypothermia reported in the literature: direct installation of cold perfusate into the epidural or intrathecal space, and intravascular cold perfusion into isolated thoracic aortic segments with the intention that the cold perfusate will be delivered through the intercostals vessels to the spinal cord. Epidural cooling for regional spinal cord hypothermia in the dog model is effective in preventing paraplegia following aortic cross-clamping.7779 Davidson et al reported a clinical trial of epidural cooling in eight patients undergoing thoracoabdominal aortic replacement for aneurysm.80 The technique satisfactorily achieved regional spinal cord hypothermia and adequate protection. Cold perfusion into isolated aortic segments has been used in animal models, with demonstration that cord temperature can be rapidly and effectively diminished.81
Intuitively, sacrifice of intercostal or lumbar arteries that are critical to the direct blood supply of the spinal cord is a significant factor in the development of postoperative paraplegia. Maintenance of flow through such arteries during all or part of the anatomical repair potentially keeps the period of spinal cord ischemia within the generally safe 30 minutes.82,83 This concept is supported by a meta-analysis of the literature reported by von Oppell in a review of 1742 patients treated for traumatic aortic rupture over a 25-year period.63 Simple aortic cross-clamping produces an incidence of paraplegia of 19.2%, whereas shunting reduces the incidence of paraplegia to 11.1%. Active augmentation of distal aortic perfusion, i.e., left atrial to femoral artery bypass or femoro-femoral bypass, has the lowest incidence of new postoperative paraplegia at 2.3% (p < .00001). The cumulative risk of paraplegia increases substantially if the duration of aortic cross-clamping exceeds 30 minutes, but only when distal perfusion is not augmented (p < .00001). In using left heart bypass for distal perfusion during replacement of descending thoracic and thoracoabdominal aortic aneurysms, Borst et al found that the technique effectively unloads the proximal circulation during aortic occlusion and maintains adequate perfusion of distal vital organs to reduce early mortality and renal failure.44 Further, the risk of spinal cord damage decreased with combined distal perfusion and aggressive reattachment of distal intercostal arteries. Clearly, the devastating complication of paraplegia and other organ failure secondary to ischemia is worthy of further research; however, a combination of measures including distal aortic perfusion, aggressive reattachment of intercostal arteries, hypothermia, avoidance of hyperglycemia, and CSF drainage has substantially reduced this devastating complication.
Motor evoked potential (MEP) monitoring has the potential to specifically reflect motor function and motor track blood supply during the period of aortic cross-clamping. MEP uses stimulation of the motor cortex or motor neurons and usually records from a peripheral muscle. In 1997, Haan et al described the technique of transcranial stimulation of the motor cortex with recording of lower extremity myogenic potentials to detect intraoperative spinal cord ischemia.84 Transcranial stimulation has not yet been approved by the Food and Drug Administration. The method requires special anesthetic techniques, since complete neuromuscular blockade is incompatible with myogenic MEP monitoring. In addition, this technique is generally used in conjunction with left atrium to femoral artery bypass. Recently, Jacobs et al published an excellent series of 184 patients undergoing TAAA repair in which they used a protocol that included left heart bypass, cerebrospinal fluid drainage, and the monitoring of MEPs. They found that MEP was a sensitive technique for the assessment of spinal cord ischemia and the identification of the segmental arteries that critically contributed to spinal cord perfusion. They were able to reduce their incidence of neurologic deficit to less than 3%.85
| RESULTS OF TREATMENT |
|---|
|
|
|---|
|
|
In patients with chronic aortic dissection, paraplegia and paraparesis developed in 3.4% versus 4.6% for patients with chronic fusiform medial degenerative disease. Consequently, the presence of chronic dissection is no longer a variable associated with the development of postoperative neurologic deficits. However, in patients with acute dissection the incidence of spinal cord ischemic sequela remains high (5 of 66; 7.6%). Reattachment of intercostal arteries was accomplished in 61.0% of the entire group of patients, but, as a result of anatomical availability, was achieved as part of the repair in 79.9% of patients with extent I and II aneurysms.
Preoperative, operative, and postoperative variables for this series were analyzed for development of postoperative neurologic deficits and early (30-day) mortality. Multivariate analysis revealed that age, rupture, symptomatic aneurysm, preoperative renal insufficiency, and total clamp time were variables predictive of early mortality, whereas rupture, diabetes, and extent II were variables predictive of paraplegia or paraparesis. In this series of 1773 patients, left heart bypass and aggressive reattachment of lower intercostal and upper lumbar arteries substantially reduced postoperative paraplegia and paraparesis.
The incidence of postoperative renal dysfunction remains a challenge. We recently reported on a group of patients undergoing Crawford extent II TAAA repair with left heart bypass randomized to renal artery perfusion of 4°C lactated Ringer's solution for renal cooling or normothermic blood perfusion from the left heart bypass circuit. Multivariate analysis confirmed that the use of cold crystalloid perfusion was independently protective against acute renal dysfunction.86
Ischemic spinal cord injury following repair remains a devastating complication. One hundred and forty-five patients undergoing extent I or II thoracoabdominal repair with a consistent strategy of moderate heparinization, permissive mild hypothermia, left heart bypass, and reattachment of intercostals arteries were randomized to cerebrospinal fluid (CSF) drainage versus no CSF drainage. In that evaluation, 9 patients (13%) in the control group developed either paraplegia or paraparesis while only 2 patients (2.6%) in the CSF drainage group developed deficits.87
A guiding principle for all patient management decisions involves determining whether the risk of the disease's natural history outweighs the risk of its treatment. In the case of TAAA repair, this must be based on each individual's risk of rupture without operation versus their risk of death or paraplegia with operation. A risk factor analysis of mortality and paraplegia after TAAA repair identified predictors of operative mortality to include preoperative renal insufficiency, increasing age, symptomatic aneurysm, and extent II aneurysm, while extent II aneurysm and diabetes were predictors of paraplegia. For patients who are acceptable candidates, contemporary surgical management provides favorable results.88
Crawford et al reported a 60% 5-year survival following TAAA replacement. Survival was reduced to 25% in patients with rupture, 40% with renal dysfunction, and 49% with coronary artery disease. The most common causes of late mortality were cardiac, pulmonary, and renal failure, sepsis, and aneurysm rupture of unoperated segments. In addition to the severe impact on lifestyle, the devastating complication of postoperative paraplegia or paraparesis decreased late survival to 44% versus 62% at 5 years.
The evaluation and treatment of patients with TAAA remain a significant challenge and require a great deal of investigative and clinical work to bring about the multifactorial approach necessary to solve the remaining complex problems.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||