Chapter 49 |
|
INTRODUCTION |
|---|
The patient population afflicted with thoracic aortic disease often harbors multiple comorbidites, including advanced age, renal insufficiency, diabetes, congestive heart disease, hypertension, and chronic obstructive lung disease. Operative intervention in this population frequently results in a significant incidence of death and long-term disability.5–7 In an effort to reduce the incidence of negative outcomes, minimally invasive techniques in the form of endovascular stenting have been introduced over the past decade.8–10 The technology, originally described by Parodi, and initially designed for use in abdominal aortic aneurysms, has been adapted for use in the thoracic aorta.11 Although these techniques were initially envisioned for use in high-risk patients, their decreased morbidity and unique applications make them suitable even for younger patients with complex abnormalities. Fortunately, attracted by the early clinical successes, industry has stepped forward with significant engineering advances over the initial "homemade" mechanical devices.12 This ongoing effort will likely permit us to expand the applicability of endovascular treatment to the various forms of thoracic aortic pathology including aneurysms and dissections as well as giant penetrating ulcers and intramural hematomas.13 The worldwide experience is growing and with this a better understanding of the indications and limitations of this innovative therapy will be elucidated.
|
DEFINITION/PATHOGENESIS/PREVALENCE |
|---|
An aortic aneurysm can be defined as an abnormal dilation of the aorta, which undergoes progressive expansion and can be fusiform or saccular in nature.14 In a Swedish study, performed between 1958 and 1985, with a population autopsy rate of 83%, the incidence of thoracic aortic aneurysm (TAA) was reported as 489 per 100,000 men and 437 per 100,000 women with a median age of 77.7 years for men and 85.3 years for women at autopsy. Approximately 5% of the lesions were thoracoabdominal aneurysms, without any difference between the sexes.15
Rupture rates have correlated with size, and a recent population-based study indicates some change in the natural history.14 In a previous report by Pressler and McNamera, the incidence was noted as 4 per 100,000 population.16 More recently, from the Olmstead County registry, the actual incidence was noted to have increased 3-fold, although the natural history was somewhat less perilous.17 Among aneurysms less than 6 cm at discovery, the rupture rate at 3 years was 16%, which increased to 31% at a size of 7 cm.18,19 Overall the incidence of rupture of thoracic aneurysms has been reported as 0.9 per 100,000 for men and 1.0 per 100,000 for women, whereas the incidence of dissection is 3.2 per 100,000 for both sexes, with fatal dissection occurring earlier in life than rupture.20
Approximately 50% of all thoracic aortic aneurysms are located in the descending aorta; these aneurysms more commonly arise at the level of the left subclavian artery and are often associated with atherosclerosis.16 Histologic examination of aneurysms reveals fragmentation and degeneration of elastic fibers in the arterial media.14 The mechanisms for this degeneration are unknown. Others have suggested that loss of structural integrity of the adventitia, not the media, is required for aneurysm formation. These pathologic changes must be differentiated from the normal aging process, in which elastic fibers fragment, smooth muscle cells diminish, collagen becomes more prominent, and ground substance increases, thus rendering the aorta less distensible and gradually more weakened.14
Heredity is thought to be involved in aortic aneurysm formation; the Yale group has demonstrated a 19% genetic predisposition to the development of familial TAA.14,20 Other risk factors for aneurysm growth include hypertension, COPD, presence of a chronic dissection, and infection of the aortic wall. Arteriosclerosis has been reported as a risk factor for aneurysms and dissections; however, many have challenged this and the general consensus is that arteriosclerosis may well be a concomitant process and not a direct cause of aneurysm formation and growth.14,20 Enlarging thoracic aneurysms can present with hoarseness due to stretching of the left recurrent laryngeal nerve, stridor from tracheal compression, dysphagia from esophogeal impingement, dyspnea from compression of the lung, and plethora or edema from compression of the superior vena cava. Patients with aneurysms of the ascending aorta may present with the signs of aortic regurgitation. Thoracic aorta aneurysms can also often be asymptomatic, detected only during work-up for another complaint. Symptomatic states, organ compression, concomitant aortic insufficiency, and acute ascending aortic dissection are widely accepted general indications for surgical intervention regardless of aortic size.18,19,20
Penetrating Atherosclerotic Ulcer
Penetrating atherosclerotic ulcer (PAU) and intramural hematoma (IMH) are distinct pathologic entities that are now being diagnosed with increasing frequency.21–28 PAU was originally described in 1934 by Shennan, who described a pathologic entity in which ulceration penetrates the internal elastic lamina into the media and is associated with a variable amount of hematoma within the aortic wall.21 It may very well be associated with aortic dissection and aneurysm formation, although it is distinct from those conditions. The ulcers are most often found in the distal descending thoracic aorta but can occur throughout the thoracic and abdominal aorta and have a characteristic appearance on computed tomography (CT) and magnetic resonance imaging (MRI).23,24
Intramural hematoma (IMH) was originally described in 1920 as a dissection without an intimal tear.24,25 The cause of IMH may be a spontaneous rupture of aortic vasa vasorum that may initiate aortic wall disintegration, eventually leading to dissection with or without an intimal tear. The term "aortic dissection without intimal flap" has frequently been applied to cases that would be classified as IMH by definition.24 Others have proposed intimal fracture of an atherosclerotic plaque as the primary event that then allows propagation of blood into the aortic media.25 Moreover, discrete penetrating atheromatous ulcers or "giant penetrating ulcers" have been proposed as a prerequisite for intramural hematoma; both aortic dissection and PAU can be accompanied by aortic wall hematoma.22,25 Regional thickening of the aortic wall greater than 7 mm in a circular or crescent shape in the absence of an intimal flap and without enhancement after contrast injection in CT and MRI is considered diagnostic of IMH.23
Aortic dissection was first described by Nicholls in 1728 and in 1826 Laennec applied the name Aneurysme Dissequant, incorrectly associating aortic dissection with aneurysms and not as a distinct entity.29 Dissection of the aorta is the more appropriate term since entry of blood into the outer two thirds of the aortic media often precedes dilation of the vessel, and dilation of the vessel may never occur. In an acute aortic dissection, the layers of the aortic wall are torn apart, creating a false lumen that runs parallel to the true lumen. The high-pressure entry site into the aortic media allows propagation of the dissection along the entire length of the aorta. Consequently, an acute aortic dissection is the most lethal of events affecting the aorta and many more people die of rupture of dissections than of aneurysms.30 The prevalence ranges from 0.2% to 0.8% of the population and usually affects males more often than females with ratios similar to that for aortic aneurysms.30
Hypertension is the single most important risk factor for thoracic aortic dissection. Aortic dissection is also a known pathology of patients with Ehlers-Danlos syndrome or Marfan syndrome. Other factors known to predispose to dissection include bicuspid aortic valve, aortic coarctation, pregnancy, and surgical manipulation of the thoracic aorta.31 Intimal tears typically occur at either the right lateral wall of the ascending aorta (type A) or just distal to the ligamentum arteriosum (type B), representing the points of presumed greatest hemodynamic stress.29
|
DIAGNOSIS |
|---|
Catheter-based angiography, in addition to being invasive and requiring nephrotoxic contrast material, effectively images only the flow lumen, and confers little information regarding the aortic wall, and no information about the mediastinum or pleural and pericardial spaces. Additionally, it requires transport to an angiography suite, and specialty personnel who may not be continuously on site. Exposure to contrast media may be minimized by use of CO2 and gadolinium.
Computed tomography (CT), although it requires radiation exposure and nephrotoxic contrast material, is usually continuously available and conveniently located. It provides detailed information about the size, location, and extent of the aortic disease, as well as involvement of the mediastinal structures and pericardial and pleural spaces. With the addition of intravenous contrast, the aorta and branch vessels can be readily imaged, and aortic true and false lumens readily distinguished.
Magnetic resonance imaging (MRI) offers many of the advantages of CT but without the necessity for ionizing radiation or nephrotoxic contrast media. However, MRI suites are frequently distant from emergency departments, frequently require more lengthy exam times, are restrictive in terms of accompanying metallic equipment, and may not be continuously available nights and weekends.
Transesophageal echocardiography (TEE), especially with multiplanar capability, offers excellent visualization of the ascending and descending thoracic aorta, cardiac structures, pericardium, and pleural spaces. Tracheal shadowing may limit visualization of the proximal and distal aortic arch, but color duplex imaging allows precise imaging of aortic wall pathology and flow between the true lumen and other areas of interest, such as across the dissection septum, or into an intramural hematoma from a penetrating ulcer.
Intravascular ultrasound, although requiring an arterial puncture and catheter laboratory imaging, provides an unparalleled view of the aortic and branch vessel intima. It can also distinguish true from false lumen, and allows definitive identification of an intimal disruption secondary to blunt aortic trauma.
|
NATURAL HISTORY AND SURGICAL OUTCOMES OF THORACIC AORTIC DISEASE |
|---|
Aneurysms of the thoracic aorta, although increasing in prevalence, are incompletely understood. The natural history is often related to the specific location and the primary cause of the disease.33–38 In patients with Marfan syndrome, dilation of the aortic root is the most penetrant of disease pathologies, and the risk of aortic dissection is directly related to the size of the aortic root.33,34
Although the size-rupture correlation is not so well established
for thoracic aneurysms as for abdominal aortic aneurysms, recent
data have better defined this relationship.18–20
Clouse et al, from the Olmstead County database, have presented
data revealing an increased incidence of thoracic aortic
aneurysms, with a 5-year risk for rupture of approximately 30%.17
Juvonen et al from Mt. Sinai in New York have identified clinical
variables that affect the risk for rupture, the most important being
increasing age, presence of COPD, maximal thoracic and abdominal
aneurysm diameter, and the presence of pain. Utilizing a
multivariable equation, the risk of rupture at one year can thus be
calculated for any patient.37
Diameters are entered in centimeters, and
pain and COPD are 0 if absent and 1 if present.
The Yale Aortic Diseases Group has documented rupture and dissection of ascending or arch aneurysms at a median size of 6 cm and descending or thoracoabdominal aneurysm ruptured or dissected at a median size of 7.2 cm.18 Overall long-term survival of patients with thoracic aneurysms at 1 and 5 years was 85% and 64%, respectively; patients with descending thoracic aneurysms had lower long-term survivals (82% at 1 year, 39% at 5 years) than did patients with ascending aneurysms (87% at 1 year, 77% at 5 years). Furthermore, the Yale group has recently published a report consisting of 1383 years of patient follow-up before surgical intervention, which permits statistically valid calculation of yearly rates of rupture or other complications.19 They found that the mean rate of rupture or dissection is 2% per year for small aneurysms, 3% for aneurysms 5.0 to 5.9 cm, and 6.9% for aneurysms of 6.0 cm in diameter or greater. The risk of rupture, dissection, or death from all causes is 6.5% at aneurysm size 5.0 to 5.9 cm and jumps to 14.1% per year for aneurysms of 6.0 cm or greater. Even more striking, when using proportional hazards regression, the odds ratio for rupture is more than 25 times higher in patients with aneurysms of 6.0 cm or greater than in those with aneurysms in the range of 4.0 to 4.9 cm.
Clouse et al reported a population-based cohort study to ascertain if the previously reported poor prognosis for individuals with thoracic aortic aneurysms had changed with better medical therapies and improved surgical techniques.17 They examined patients with the diagnosis of degenerative thoracic aortic aneurysms among the residents of Olmsted County, Minnesota, between 1980 and 1994 and compared them with a previously reported cohort of similar patients between 1951 and 1980. The cumulative risk of rupture was 20% after 5 years, with 79% of ruptures occurring in women. The interval between initial diagnosis and rupture was 4.3 years, with rupture being the leading cause of death. Of the variables examined in univariate analyses, eventual development of dissection within the aneurysm, female sex, symptoms at diagnosis, and age at diagnosis were related significantly to aneurysm rupture, whereas smoking, chronic obstructive lung disease, hyperlipidemia, family history, and saccular configuration were not. Importantly, the overall 5-year survival improved to 56% between 1980 and 1994 compared with only 19% between 1951 and 1980.
In reviewing the outcomes of those patients with thoracic aneurysm who do not undergo surgery, it has been found that 50% to 60% succumb to rupture.39 By way of comparison, of all those who undergo elective surgery, 70% are alive at 2 years, and 59% are alive at 5 years, which further emphasizes that the natural course of thoracic aortic disease can be favorably altered by appropriate surgical intervention.40
Extensive experience has revealed that in patients undergoing thoracoabdominal repairs for aneurysm disease, the survival rate can reach 92% at 30 days and 60% at 5 years.40–42 The overall incidence of paraplegia and paraparesis ranges from 3% to 16% and the significant predictors were total aortic clamp time, extent of aorta repaired, aortic rupture, patient age, proximal aortic aneurysm, and history of renal dysfunction.43–45 Despite the advances in surgery, the incidence and attendant morbidity and mortality of postoperative acute renal failure (ARF) have not declined substantially in patients undergoing thoracic aorta surgery. For patients with extensive aortic disease, the risk of hemodialysis is 5% for those with normal preoperative renal function and 17% for those patients with preoperative renal dysfunction.46
Traumatic rupture and resultant pseudoaneurysm of the aorta remains a surgical challenge because it is a diagnosis that is not easily made and is frequently associated with other serious injuries, and its repair is associated with high morbidity and mortality.47 More contemporary theories propose that traumatic aortic rupture is a complex multivariate process secondary to a combination of stresses.48 The mortality of this injury is in relation to the seriousness of the lesion, associated visceral injury, and the great urgency of the operation. The morbidity is in relation to the difficulty of spinal cord protection, which leads to paraplegia in young patients.47,48
Penetrating Atherosclerotic Ulcer
The clinical presentation of penetrating atherosclerotic ulcer (PAU) is similar to that of classic aortic dissection,21 but very little is known of the natural history of PAU since most patients in the literature have undergone surgical therapy or have had only short-term follow-up. Recently, more follow-up has become available as this disease entity has become better understood.22 The risk of aortic rupture is higher among patients with PAU (40%) than with patients with type A (7.0%) or type B (3.6%) aortic dissection.25 However, the progression of PAU is slow and is associated with a low incidence of acute rupture or other life-threatening events. Among patients with PAU who are not treated surgically, the natural history would appear to be that the majority will have enlargement with the formation of saccular or fusiform pseudoaneurysms and intramural thrombus.22 Age and general health of the patient, the location of the PAU, and the rate of growth should therefore all be considered when deciding whether or when resection is appropriate.22,26
Aggressive control of blood pressure in an ICU setting has been the initial management of intramural hematoma (IMH).24 The overall mortality of IMH has been shown to be high in the first month after acute onset of symptoms and particularly for IMH of the ascending aorta. Furthermore, the development of mediastinal hemorrhage and pericardial and pleural effusion has been found to be more frequent in patients with aortic IMH than in patients with aortic dissections.23 The highest mortality rate has been reported among patients in whom the hematoma originated in the ascending aorta and aortic arch. Also, patients with severe artherosclerosis of the ascending aorta and arch have a high prevalence of atheromatous emboli in the cerebral circulation.25 Experience has therefore suggested that a more aggressive approach with early surgery is warranted in patients who have ascending aortic involvement or in those who have a coexisting aneurysm with IMH.23,24,27
For patients with Stanford type A dissections, surgical intervention is performed immediately after diagnosis to avert the high risk of death due to various complications such as cardiac tamponade, aortic regurgitation, and myocardial infarction.49–51 Untreated, type A dissections are associated with a mortality rate of 1% to 2% per hour during the first 24 to 48 hours. The Stanford experience revealed that the operative mortality for acute type A dissections was 7%; for chronic type A dissections, the operative mortality was 11%. Overall, the 5-year actuarial survival rate for discharged patients was 78% for type A dissections.52–54
In contrast, the preferred treatment for most patients with Stanford type B (descending aorta) is less well defined.53–55 Although anti-impulse therapy, with beta blockade as the primary therapy, has been the most accepted therapy, it is still associated with a 10% to 15% hospital mortality rate, with significant morbidity in the long-term management from aneurysmal dilation of the false lumen.55 Surgical treatment has been reserved for specific complications, namely, intractable pain, impending rupture, malperfusion syndromes, and early expansion to a diameter greater than 5 cm.29,54,56 Other conditions that should prompt consideration of early operation include Marfan syndrome, presence of sizable localized false aneurysm in the proximal descending thoracic aorta, arch involvement, and expectations of poor medical compliance.52,55 For this cohort of complicated patients, however, surgical procedures have been associated with mortality rates as high as 60% to 70%, primarily because of the unpredictable effect a central aortic operation may have on more peripheral perfusion. The actuarial survival for all patients with type B dissection was found to be 65% at 1 year and 50% at 5 years. More specifically, survival was 73% at 1 year and 58% at 5 years for medically treated patients, and 47% at 1 year and 28% at 5 years for surgically treated patients.53 However, these are not comparable cohorts, as surgical patients are self-selected with their complications.14,31,53
|
ENDOVASCULAR DEVELOPMENT AND THERAPY |
|---|
|
Some new terminology must be introduced in relation to this technology.12 Endoleaks, the failure to exclude the aneurysm sac from the blood stream, are classified as types I to IV. Type I endoleaks occur at the proximal or distal attachment sites, and signify a failure to achieve a hemostatic seal at these implantation sites.9,57–61 Type II endoleaks denote a communication between a branch vessel and the excluded aneurysm sac. These usually occur from a back-bleeding inferior mesenteric artery in the abdomen, or intercostal arteries in the chest. Both an entry and exit vessel are usually necessary for prolonged patency. Type III endoleaks originate from the mid graft sections, and are usually caused by disruption of graft-to-graft overlaps, or by leakage through the graft itself. Finally, Type IV endoleaks are characterized by an increase in size of the aneurysm sac in the absence of an identifiable patent branch vessel, variously ascribed to "endotension." Regardless of type, any endoleak associated with expansion of the aneurysm sac implies a procedural failure.
Descending Thoracic Aortic Aneurysm
The anatomical features of descending thoracic aneurysm make them an almost ideal substrate for this technology. Since they frequently arise distal to the left subclavian artery, and proximal to the celiac axis, adequate landing zones (at least 2 cm in length) are frequently available, usually within relatively straight segments, and with no critical side branches. Femoral and iliac arteries greater than 8 mm in diameter were necessary to allow introduction of a 28F (OD) delivery sheath in Stanford patients. Devices were limited to a maximal diameter of 40 mm; aortas larger than 37 mm in diameter were themselves likely to be aneurysmal, and unlikely to serve as a stable attachment zone. Other anatomical constraints that precluded secure fixation included acute angulation at the distal arch, and severe sigmoid-like tortuosity coursing through the diaphragmatic crura, reflecting the relative inflexibility of these crude first-generation devices.58–61 Commercially produced second-generation stent grafts (Fig. 49-2) are more flexible, and have a lower profile than these earlier devices.
|
|
|
Patients with multilevel aortic disease can present a formidable challenge for the cardiothoracic surgeon.38,64 The morbidity rate of surgical repair can be substantial in these patients, and it is frequently compounded by the requisite second operation. In patients with an abdominal aortic aneurysm, 5% also have a descending thoracic aneurysm; in patients with a descending thoracic aneurysm, 13% to 29% also have abdominal involvement.35,38 Sequential repair requires an interval period for recovery during which time the second aneurysm can rupture. Crawford found that 30% of early postoperative deaths after isolated repair of a descending thoracic aneurysm were caused by rupture of an untreated infrarenal aneurysm. In asymptomatic patients, Crawford recommended repair of the thoracic aorta initially.33 While awaiting the second operation, the abdominal aneurysm is easier to observe, and if symptoms or rupture occur, patients could potentially survive for periods sufficient to attempt emergency repair. In contrast, thoracic aneurysms are more likely to rupture without warning and are rapidly fatal within minutes. In the Stanford study, simultaneous repair was successfully undertaken in 17 patients.61 Utilizing a retroperitoneal approach to the abdominal aorta, an endovascular repair of a thoracic aortic aneurysm could be accomplished through an 8-mm side limb attached to the abdominal graft. A thoracotomy was therefore avoided, and complete aneurysm exclusion was achieved in 94% of patients. One patient died, resulting in a hospital mortality rate of 6%. This may prove to be a valuable addition to the surgical armamentarium for patients with both abdominal and thoracic aortic aneurysms.
Penetrating Atherosclerotic Ulcer
Penetrating ulcers present perhaps one of the most appealing clinical indications for this stent graft technology.65 Presenting in an elderly population with extensive comorbidities, these diffusely diseased aortas present significant challenges for conventional repair. Poor tissue integrity combined with a high likelihood for intraoperative thromboembolism is a prime setting for severe complications.21,22 Additionally, because of the diffuse nature of this process, representing end-stage atherosclerotic disease, it is difficult to limit the extent of resection and consequently the likelihood of complications (Fig. 49-5). Simple stent graft coverage of the penetrating ulcer can limit the progression of IMH and allow healing to occur. Unfortunately, even with successful stent graft implantation, retrograde aortic dissections and new ulcer formation have been noted in a significant percentage of patients, amplifying the diffuse and severe nature of this disease.65
|
Radiographically, intramural hematomas may be indistinguishable from an acute dissection with a thrombosed false lumen, and in fact may represent one end of a spectrum of disease, with IMH and intact intima at one extreme, and full-blown aortic dissection at the other. Although pure IMH of the thoracic aorta is not amenable to stent graft repair, any secondary intimal defect could be covered, perhaps limiting progression of the disease. IMH of the ascending aorta is problematic, as progression to a classical type A dissection occurs in a significant percentage of patients.21,25
Aortic dissections may present the greatest challenge, but also the greatest utility, for thoracic stent grafts.66–78 For Stanford type A dissections, with involvement of the ascending aorta, open surgical repair is indicated. As more aggressive repairs of acute type A dissections are undertaken in an effort to minimize late aneurysmal complications, as advocated by Kazui et al, stent grafts may play a role.78 In younger patients, or those afflicted with Marfan disease, aneurysmal complications are more likely to develop late in the time course of a chronic dissection. Aneurysmal changes of the transverse arch are particularly problematic. In an effort to minimize these late complications, Kazui has advocated ascending and arch replacement for these younger, good-risk patients, with very acceptable mortality. Subsequent aneurysmal degenerative changes would be limited to the descending and abdominal aorta, more easily manageable if the arch is already repaired. Some authors have taken this argument one step further, and advocated insertion of a stent graft into the true lumen of the proximal descending thoracic aorta both to minimize late aneurysm formation as well as to promote false lumen thrombosis, potentially avoiding a double lumen aorta.79 Although such an approach may seem unusually aggressive, if a particular population at risk could be defined, stent graft placement could prevent the late development of aneurysmal complications in a high-risk group.
For Stanford type B dissections, however, this stent graft technology offers many new advantages. Currently, medical management utilizing anti-impulse therapy is the most widely applied therapy for uncomplicated type B dissections. Surgical repair has been reserved for patients presenting with complications, including intractable pain, rapid expansion to a diameter greater than 4.5 to 5 cm, malperfusion syndromes, and leak or impending rupture.29–31,54 Unfortunately, in this setting, surgical mortality may exceed 60% to 70%.30 With medical management, which itself carries a 10% hospital mortality, it is estimated that approximately 70% of these patients will have a persistently patent false lumen, and 20% of these will become aneurysmal.30,55
The application of this stent graft technology could profoundly affect the natural history of aortic dissections (Fig. 49-6). Coverage of the primary intimal tear redirects flow into the true lumen, which will not only provide protection from rupture, but will also eliminate malperfusion resulting from dynamic obstruction ("true lumen collapse") in nearly all patients.74 False lumen filling, with the attendant risk for rupture of its aneurysmal dilation, is eliminated, and false lumen thrombosis results in a single-barrel aorta.80 If the procedure is performed in the catheterization laboratory, subsequent catheter-based investigation can reveal continued malperfusion secondary to uncorrected static obstruction, which can then be expeditiously addressed with uncovered orifice stents into the true lumen.74 This is clearly a more reliable procedure for restoring visceral perfusion than a central aortic operation, and a less morbid procedure than a direct operation on the visceral arteries themselves. Moreover, the stent graft need not be overly long, rarely extending distal to T-6, thus minimizing the risk for paraplegia. In a previously reported series, 12 of 15 complicated type B dissections were successfully repaired with this approach, with an operative mortality of only 20%, and restoration of a single-lumen aorta in 70% to 80% of patients.74
|
However, one must not discount the possibility for iatrogenic injury during instrumentation for stent graft placement. In the acute phase, the outer layer of the false lumen is a thin and friable adventitia, and its perforation, perhaps even with just a guidewire, would be catastrophic. Similarly, the dissection septum is also quite fragile, and may provide an insufficient distal anchor for a stent graft. There are some examples (Dake; personal communication) of septal perforations at the distal stent graft fixation site, suggesting the susceptibility of the acute dissection septum to repetitive erosive trauma. Even proximal secure placement within a nondissected aorta may pose difficulties, as there have been reports of false aneurysm formation in the transverse arch after placement of an endograft with its proximal portion uncovered. The primary intimal tear may also be quite extensive, and extend proximally to the left subclavian artery orifice. Effective exclusion of false lumen perfusion in these instances may be difficult to achieve, requiring implanting the stent graft just distal to the left carotid artery, and covering the left subclavian orifice. Although it has been suggested that the left subclavian artery may be sacrificed with impunity, the Stanford approach has been to create a left carotid to subclavian transposition or bypass, and then ligate the left subclavian proximal to the vertebral artery takeoff so as to eliminate the possibility of retrograde filling of the aneurysm sac from the left subclavian stump.57,74,76,77 Another attractive possibility would be that of a branched stent graft for the left subclavian, or even all the arch vessels as advocated by Inoue of Japan.84
Many unanswered questions remain. For chronic dissections, and for thickening and fibrosis of the septum, the origin of critical branch vessels from both true and false lumens, as well as the presence of multiple true lumen to false lumen communications, would seemingly markedly limit the utility of this stent graft technology. The presence of critical side branches, especially low intercostal arteries arising from the false lumen, or the presence of critical visceral arteries has tempered enthusiasm for endograft insertion into the true lumen because of the risk for ischemic injury. However, should the motivation for stent graft repair become compelling, stent grafting into the true lumen could be performed, and then false lumen flow could be assured or augmented by angioplasty and stenting of septal perforations. There have been some short-term successes in managing aneurysmal expansion of the false lumen. The proximal portion of the descending thoracic aorta seems inordinately disposed toward false lumen dilation, and stent grafts can frequently be placed to cover the primary entry tear in that aortic segment. Although septal fenestrations usually persist at the level of the diaphragm, as well as more distally around the visceral vessel orifices, thrombosis of a proximal aneurysmal dilation of the false lumen has been achieved in a small number of cases. Whether this thrombosis will remain stable over the long term is unknown.
Ten years of experience with endovascular AAA repair has yielded important information regarding the relationship between stent graft design and stent graft performance.85–88 For example, tapered, flexible, over-the-wire delivery systems (less than 20F in diameter) rarely fail to traverse tortuous iliac arteries. Hooks at the proximal stent graft appear to provide the most secure means of proximal attachment; column strength is of little value. Although modular stent grafts are more versatile than unibody stent grafts, graft-to-graft attachments and overlaps produce unusual stresses and are similarly prone to late failure. Severe stresses are exerted on thoracic grafts at areas of angulation, predisposing to fatigue fractures. Any movement between the stent body and the overlying fabric will lead to graft erosion. Most important is the absolute necessity for long-term follow-up. From the Eurostar registry, up to 10% of patients per year may require secondary procedures to assure exclusion of the aneurysm sac.85 The evolution of the aneurysm sac is a dynamic process that requires monitoring over years. Failure of the sac to stabilize and/or shrink in size suggests that protection from rupture has not been conferred. Longitudinal follow-up is essential, both to survey the fixation sites and to monitor the aneurysm sac. Other limitations include anatomic constraints, such as the need for adequate proximal and distal aortic necks, the confounding presence of critical aortic branches in the diseased aorta, and the fact that only relatively straight segments of aorta can be managed in this manner. Placement of the stent graft close to the distal arch appears to be associated with a higher incidence of strokes, presumably due to catheter manipulation in the ascending aorta and arch. Even guidewire manipulation within the severely atherosclerotic transverse arch probably poses some finite risk for atheroembolism and stroke.
There are currently no commercially available devices in the United States. Devices under evaluation include the Medtronic Talent, a new iteration of the Gore Excluder, and likely other thoracic stent grafts from the major abdominal stent graft manufacturers. Many of these will likely appear in Europe. In Japan, where this technology has been widely embraced, the vast majority of stent grafts are individually constructed at various centers. The improvements in second- and third-generation devices will likely supplant these devices as they become available. Already, these include lower profile, greater flexibility, increased conformability, and greater ease of insertion and deployment. Whether the perceived advantages of reduced operative time, blood loss, hospital stay, and overall complications and mortality will be confirmed awaits further study. Interestingly, the phase II FDA study of the Gore Excluder included a surgical control arm. Although this was not a randomized surgical cohort, it is the first study to compare results in similar patient populations.
Several authors have reported results of elective stent grafting for Type B aortic dissections using the Talent graft, the Excluder graft, and homemade graft (Fig. 49-6).63,65,69 Complete coverage of the primary tear was achieved in most instances, and thrombosis of the false lumen was achieved in 70% to 80% of cases. Few complications were noted with the exception of the "postimplantation syndrome," which usually resolved within 4 to 7 days. However, although it is clearly possible to treat type B dissections, more problematic is which patients should be treated. Only further experience, with clear understanding of potential complications and demonstration of long-term durability, will clarify this dilemma.
There is another subset of dissections for which this stent graft technology may have particular utility. For approximately 5% of patients with type A dissections, the primary tear is distal to the left subclavian, with retrograde propagation into the ascending aorta.78,79 These patients can be quite challenging, as conventional ascending aortic repair leaves the primary intimal tear untouched. Alternatively, proximal descending replacement does not address the dissected ascending aorta. Stent graft coverage of the primary intimal tear has allowed healing of both the ascending and descending aorta.78,79
Von Segesser et al have proposed new strategies for the treatment of dissection of the descending thoracic aorta extending back into the ascending aorta such that the dissection should be treated in accordance with the site of the predominant lesion (Fig. 49-7).89 Replacement of the arch with a variable portion of ascending aorta via median sternotomy is recommended in patients with an enlarged aortic diameter, pericardial effusion, and/or aortic insufficiency. Predominantly distal dissections with a dilated descending thoracic aorta, distal complications, or both are best approached via lateral thoracotomy. Predominantly distal dissection with an almost intact, relatively small ascending aorta, a small aortic arch, and an already thrombosed false ascending aortic lumen is treated in the same manner as type B aortic dissection, in which stent graft repair is an alternative to conventional surgery. Kato et al recently described their experience in 10 type A aortic dissections in which the entry tears were located in the descending thoracic aorta in all cases.79 Z-stents were placed in the true lumen of the proximal descending thoracic aorta in 9 of 10 patients and in the mid descending thoracic aorta in the remaining patient. Complete closure of the entry tear was achieved at the end of the procedure in all patients and complete thrombosis of the false lumen of the ascending aorta was observed after stent grafting in all patients. No procedure-related complications were observed and, after a mean follow-up of 20 months, no aortic rupture or aneurysm formation was noted. They therefore concluded that stent graft repair of aortic dissection with an entry tear in the descending thoracic aorta is a safe and effective method and may be an alternative to surgical graft replacement in highly selected patients.79
|
|
CONCLUSIONS AND RECOMMENDATIONS |
|---|
It should be emphasized that these patients with thoracic aortic disease represent multidisciplinary efforts involving both surgeons and radiologists. Stent graft placement requires state of the art imaging technology as well as precise manipulation of catheters and positioning of the stent. Aortic dissection remains a surgical disease; even in the context of catheter-based treatment, the management of aortic dissection and its complications requires seasoned judgment and, in some cases, surgical intervention on an urgent basis. Accordingly, surgical specialists must remain involved in the treatment of aortic dissection, irrespective of the type of treatment.
|
REFERENCES |
|---|
