- Email: [email protected]
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
~~
~
~
07334651199 $8.00
DISEASES OF THE AORTA
+ .OO
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS Penetrating Atherosclerotic Ulcers and Intramural Hematomas Michael A. Coady, MD, MPH, John A. Rizzo, PhD, and John A. Elefteriades, MD
Penetrating atherosclerotic ulcer (PAU) and intramural hematoma (IMH) of the aorta were virtually unknown in the prior era of aortic imaging by aortography. These disorders were not defined as radiographically dis33 In the current tinct until the mid-1980~.~~, era of three-dimensional, high resolution imaging of the aorta by computerized tomography (CT), magnetic resonance (MR) imaging, and transesophageal echocardiography (TEE), these two disorders have become increasingly recognized. The distinctions from typical aortic dissection have not always been clear in terms of diagnosis, clinical characteristics, and treatment. Most reports of PAU and IMH in the literature have concentrated on identifying these disorders as entities distinct from classic aortic dissection by describing small series of these patients.* There is growing concern that the anatomic pathology and clinical behavior of PAU and IMH may differ substantially from those of typical aortic dissection, and that clinical management may need to be spe*References 1, 4, ll, 14, 15, 19, 23, 28
cifically tailored. In particular, it is becoming more apparent that the paucity of information and relative rarity of diagnosis are due to the fact that cases of PAU and IMH are often misdiagnosed and managed as "aortic dissections." The infrequence, misdiagnosis by initial radiologic inspection, and lack of clinical experience with PAU and IMH has not permitted definition of optimal therapy for these two variants of aortic disease. The present review draws on our experience at the Yale Center for Thoracic Aortic Disease in order to clarify the disorders of PAU and IMH of the aorta. This article expands on our initial report2 and reviews the clinical features, radiographic appearance, pathologic findings, and natural history of individuals with PAU and IMH. Appropriate correlates regarding the optimal treatment of these individuals are also discussed. CLASSIC AORTIC DISSECTION
A classic aortic dissection begins with a laceration of the aortic intima and inner layer
From the Department of Surgery, Section of Cardiothoracic Surgery (MAC, JAE), and the Department of Epidemiology and Public Health (JAR), Yale University School of Medicine, New Haven, Connecticut
CARDIOLOGY CLINICS OF NORTH AMERICA VOLUME 17 * NUMBER 4 NOVEMBER 1999
637
638
COADY et a1
of the aortic media, which allows blood to course freely along a false lumen in the outer third of the media (Color Fig. 1). An artist’s rendition of a typical aortic dissection is presented in Fig. 2, and PAU and IMH are illustrated for later comparison. As is easily appreciated, a classic aortic dissection involves a dissection flap that traverses the aortic lumen, dividing the aorta into true and false lumina. The artist’s rendition corresponds well to the operative findings familiar to the aortic surgeon. The flap traverses the aortic lumen obliquely in cross-sectional views; it is not oriented circurnferentially. This standard appearance of classic aortic dissection will later be contrasted with imaging findings in PAU and IMH in our patients. Understanding the pathogenesis of aortic dissection requires consideration of the inciting event causing the propagation of blood within the aortic media. Although various risk factors predisposing the aorta to dissection have been well-described, the precise insult leading to laceration of the intima and media remains unclear. Historically, the primary etiologic event leading to aortic dissection has been extremely controversial. Cystic medial necrosis associated with Marfan’s disease and other connective tissue disorders was once believed to contribute to aortic medial degeneration, leading to aortic dissection. Larson and Edw a r d ~ , ’however, ~ demonstrated that only a few of their 161 patients with known aortic dissection exhibited medial degeneration. They found that 158 of these autopsies had intimal aortic tears, supporting the theory initially proposed by Murray and Edwards,I6 that the intimal tear is the primary event,
allowing the blood to spread through the outer two-thirds of the aortic media. Cystic medial necrosis is no longer regarded as the common structural disorder underlying aortic dissection. Gore6in 1952 suggested that the etiology is related to spontaneous rupture of the aortic vasa vasorum leading to subsequent intimal tear and aortic dissection. He accumulated 29 cases of aortic dissection without intimal tear, describing a process in which IMH precedes intimal rupture because hemorrhage of the vasa vasorum weakens the media, allowing blood to leak into the media from the aortic lumen owing to arterial pressure. Degenerative changes within the media itself can affect its elastic constituents. Loss of these elements reduces the resistance of the aortic wall to hemodynamic stress, leading to aneurysmal dilation of the vessel with subsequent dissection. Our group at the Yale Center for Thoracic Aortic Disease has recently confirmed the increasing incidence of aortic dissection with greater aortic diameters (Figure 3).3In the dilated aorta, the stress leading to aortic dissection is composed of a radial and a longitudinal vector. As the aorta becomes increasingly more aneurysmal, the longitudinal vector becomes the more significant force. This explains the typical occurrence of a transverse tear along the course of the aortic wall. Inheritable disorders of elastic tissues also predispose to the development of aortic dissections, including Marfan,22Turner,27Nooand Ehlers-Danlos S y n d r ~ m e sOf .~~ these Marfan Syndrome is the most common and best understood. Other known risk factors for the development of aortic dissection
Figure 1. Classic aortic dissection. Hematoxylin-eosin stain of aortic wall (original magnification x 100). Note the medial disruption with blood seen coursing through the outer two-thirds of the aortic media. Figure 15. Aortic wall (hematoxylin-eosin, original magnification x 100) of a patient with rupture of a penetrating atherosclerotic ulcer. The aortic intima has been replaced with cholesterol clefts that have burrowed though the aortic media. Blood is seen on the outside of the media. There is no remaining adventitial layer. Figure 16. Aortic wall (hematoxylin-eosin, original magnification x 100) of a patient with an intramural hematoma. Note the lack of intimal disruption and medial disruption from blood within this layer.
639
640
COADY et a1
Figure 2. Aortic dissection, intramural hematoma, and penetrating atherosclerotic ulcer. Note the distinct intimal flap separating the true and false lumens in aortic dissection. An intramural hematoma, possibly caused by rupture of the vasa vasorum, leads to a concentric hematoma within the media, and has no intimal tear. Penetrating atherosclerotic ulcers burrow deeply through the intima into the aortic media, precipitating a localized intramedial dissection associated with a variable amount of hematoma.
include pregnancyz4and inflammatory diseases of the media.5 Significant derangement and loss of elastic tissue are common in younger patients with aortic dissection, especially those with congenital defects; however this correlation in older patients has not held welL9Some investigators believe that a loss of smooth muscle cells, often appearing in a laminar pattern, is the predominant factor leading to aortic dissection in patients older than 40 years.7 Because degenerative changes do not fully explain the occurrence of the dissection, various mechanical forces have also been considered. These include flexional forces of the ves-
sel at fixed sites, the radial impact of the pressure pulse, and the shear-stress of the blood. The heart, ascending aorta, and aortic arch are relatively mobile, whereas the descending aorta is suspended to the spinal column. During the cardiac cycle, the heart and aorta produce rhythmic movements, allowing all but fixed segments to move.31These fixed points are exposed to the most significant flexional forces. Classic type A and B aortic dissections may produce an intimal tear at the areas of greatest hydraulic stress: the right lateral wall of the ascending aorta, or the descending thoracic aorta in proximity to the ligamentum arteriosum. These are the most
(6/7 1) 8.5% (4/47) 8.5%
( 12/42) 28.6%
(1/14)
7.1%
n -
5
26
Thoracic Aortic Size (a) Figure 3. Increasing incidence of aortic dissection with larger aortic sizes. (From Coady MA, Rizzo JA, Hammond GL, et al: What is the appropriate size criteria for resection of thoracic aortic aneurysms? J Thorac Cardiovasc Surg 113:476491, 1997.
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
641
Figure 4. Computerized tomography (CT) of a patient with a classic type B aortic dissection. Note the intimal flap and presence of both true and false lumina.
common sites of aortic dissection, emphasizing the significant impact that flexional forces have on the aortic wall. Hypertension and increased aortic blood pressure alters the slope of the pulse pressure wave, double-pole double-throw (dp / dt), and has been shown to add a mechanical strain on the aortic wall leading to deposition of additional elastic lamellae.8,32 There is also an increased propensity for delamination of the aortic wall, because of the shearing forces exerting a longitudinal stress along the aortic wall. At one time it was also thought that atherosclerosis was the cause of acute aortic dissections. Among the earliest believers in this theory were Virchow (1851), Girode (1887), Ewald (1890), Rolleston (1893), and von Moller (1906).26In a classic paper entitled ”Dissecting Aneurysms,” Shennan in 1934 began to recognize that although nodular atheromas were present in a large number aortic dissections and ruptures, only a few displayed a relationship of the atheroma to the exact location of primary rupture or dissectionZ6Only 6 of 218 dissections in his study demonstrated evidence of a dissection at the base of an atheromatous ulcer. There still remains some controversy as to whether atherosclerosis causes aortic dissection.’” Most authorities believe it does not.
Those who believe that atherosclerosis causes aortic dissection do not reason that aortic dissections are more frequently found in the ascending aorta, where atherosclerosis is less common. Additionally, dissection in an area of gross atherosclerosis is usually limited, because of neighboring fibrosis and calcification. The imaging studies useful in diagnosis classic aortic dissections demonstrate two or more channels separated by an intimal flap on CT scans (Fig. 4), ECHO (Fig. 5), and angiography (Fig. 6). These images are presented to permit later comparison with findings in PAU and IMH. PENETRATING ATHEROSCLEROTIC ULCERS
In 1934, Shermanz6was the first to describe penetrating atheromatous ulcers of the thoracic aorta. These atheromatous plaques, characterized further by Stanson et aIz8in 1986, ulcerate and disrupt the internal elastic lamina, burrowing deeply through the intima into the aortic media.4 The plaque may precipitate a localized intramedial dissection associated with a variable amount of hematoma within the aortic wall, may break through into the adventitia forming a pseudoaneurysm,
642
COADY et a1
Figure 5. A, Transesophageal echocardiogram (TEE) showing a transverse plane at the level of the diaphragm showing a dissection flap. B, TEE showing a long axis plane at the level of the aortic valve, showing an extensive aortic dissection anteriorly and posteriorly. LA = left atrium.
or may rupture completeIy into the right or left hemithorax. Ulceration of an aortic atheroma occurs in patients with advanced atherosclerosis; however, this process is usually asymptomatic, confined to the intimal layer, and not associated with an intramural hematoma. When an atherosclerotic plaque penetrates into the media, the media is exposed to pulsatile atrial flow, causing hemorrhage into the wall without an intimal flap. This PAU involves a localized dissection, and the intramural hematoma is prevented from extension by surrounding transmural inflammation and relative fusion of the aortic wall layers.21The adventitial layer is frequently dissected from the media in patients who have penetration through the media.30The adventitia may also rupture, and
only the surrounding mediastinal tissues then contain the hematoma. Patients with PAU, therefore, have a localized dissection which is limited by areas of severe calcification associated with locally advanced atherosclerotic disease, therefore representing a different aortopathy than that of a classic aortic dissection. The site of entry into the dissection in patients with penetrating ulcers is the ulcer itself. On postmortem examination, there may be several atherosclerotic ulcers, often extending into the adventitia. The clinical presentation of patients with penetrating aortic ulceration is similar to that of classic aortic dissection; however, penetrating ulceration represents a unique pathologic event that may have distinct prognostic and therapeutic implication^.'^ Since preliminary reports4,14, ** indicate that penetrating ulcer is distinct from aortic dissection, our group recently sought to isolate true cases and study the natural history of these atheromatous ulcers. Because penetrating ulcer is much less common than classic aortic dissection, it may be often unrecognized, and therefore misclassified as an aortic dissection on presentation. Penetrating ulcers present similarly to typical aortic dissection, with chest or back pain of excruciating intensity. Absent, however, are findings of branch vessel occlusion or run-off organ ischemia, as PAU does not generally impair luminal blood flow. On imaging by CT, MR, and TEE, the cardinal findings for PAU include an ukev crater similar to the barium appearance of a duodenal ulcer. In the artist’s rendition (see Fig. 1) of PAU and IMH corresponds closely with the images engendered by their verbal designations. PAU is shown as an ulcer crater with varying degrees of surrounding hemorrhage. PAU does not have a flap traversing the aortic contour and defining a true and false lumen. These renditions corresponds well to current understanding of the pathophysiology of these disorders. The diagnosis of PAU is made on CT scan and angiography by demonstration of a contrast-filled outpouching in the aorta in the absence of a dissection flap or a false lumen, and often in the presence of extensive aortic calcification (Fig. 7 and 8).11TEE demonstrates a craterlike ulcer with jagged edges, usually
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
Figure 6. Angiography illustrating a classic aortic dissection. Note the dissection flap indicated by the arrows.
Figure 7. CT scan of a patient with a penetrating atherosclerotic ulcer at the level of the mid-descending thoracic aorta. Clearly demonstrated is a contrast filled ulcer. No intimal flap is seen.
643
644
COADY et a1
Figure 8. Angiography demonstrating a penetrating atherosclerotic ulcer.
in the presence of extensive aortic atheroma; there is absence of an intimal flap or a false lumen (Fig. 9). On MR imaging scans (Fig. lo), PAU are diagnosed by the visualization of a distinct ulcer crater in the absence of an intimal flap or false lumen. In order to confirm the existence of penetrating ulcers, a recent retrospective review of every patient admitted with a preliminary diagnosis of aortic dissection at Yale-New Haven Hospital from 1985 to 1998 was conducted.2Of 212 patients initially diagnosed as having an aortic dissection, 19 individuals (9%) (9 males and 10 females) were found to have PAU on re-review of CT, MR imaging, angiography, ECHO imaging studies, intraoperative findings, or pathology results. Demographic information, complications, age profiles, aortic sizes, location, hospital course, and follow-up were compared with classic dissections and intramural hematomas. INTRAMURAL HEMATOMAS (IMH) Although classical aortic dissections begin with an intimal tear which allows blood to
course rapidly along a plane in the outer third of an intrinsically diseased media, an IMH is thought to begin with rupture of the vasa vasorum of the aortic wall. This results in a circumferentially oriented blood-containing space seen on three-dimensional imaging studies (see Fig. 1). Note in Figure 1 that IMH is shown as a blood-filled circumferential space without ulcer or other intimal break or discontinuity. Krukenberg first described IMH in 1920 as a ”dissection without intimal tear.”19 This variation of aortic dissection was considered a distinct entity at postmortem examination. On histologic analysis, a hematoma is visualized disrupting the medial layer of the aorta. In 1985, Yamada et al,33 proposed the term ”aortic dissection without intimal rupture” to describe intramural hematomas in isolation of intimal disruption or penetrating ulceration. In their series, 6 of the 14 patients (43%) had involvement of the ascending aorta, and the remaining 8 patients (57%) had involvement the descending aorta. All patients were medically managed with serial imaging studies; 3 patients (21%) died within 1 month of diagnosis (2 ascending and 1 descending) and another died following a graft replacement. Nine of the 10 survivors showed complete resolution of the aortic intramural hematoma within 1 year. IMH may occur spontaneously in hypertensive patients, or following blunt chest trauma.’ It has been speculated that aortic intramural hematoma may also originate from ulceration in an atherosclerotic aorta; therefore, the cause may be m~ltifactorial.’~ These individuals may have no evidence of an atheromatous ulcer on imaging studies, however, and an intimal tear is not visualized. Because there is no intimal discontinuity, the space does not communicate directly with the aortic lumen and, unlike the false lumen of classic aortic dissection, does not generally show enhancement with contrast administration on CT scanning (Fig. 11) and angiography.19 On TEE examinations, IMH displays an aortic wall hematoma in the absence of an intimal dissection flap or penetrating ulceration (Fig. 12).
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
645
Figure 9. TEE demonstrating a penetrating atherosclerotic ulcer.
THE YALE EXPERIENCE: DISTINGUISHING PAU AND IMH FROM CLASSIC AORTIC DISSECTION A total of 529 imaging studies (CT, MR imaging, angiography, and TEES) were carefully re-reviewed on 214 patients originally classified as having an aortic dissection (87 type A and 127 type B). These studies were scrutinized with a radiologist specifically trained in vascular imaging. Among these patients, 36 were found on re-review to meet radiographic or morphologic criteria (described below) for the aortic dissections variants of PAU (19 patients) or IMH (17 patients). These 19 patients with PAU and 17 patients with IMH were then compared statistically to the remaining patients with classic aortic dissections (80 type A and 98 type B remaining after re-review).
Imaging Results The re-review of imaging studies initially classified as aortic dissection did indeed confirm the existence of radiologically distinct entities of PAU and IMH. Of the 212 patients
with an initial diagnosis of an aortic dissection, 36 patients met the criteria for clinical variants-19 with PAU and 17 with IMH. PAU and IMH were found to be more focal lesions than classic aortic dissection, which frequently propagates for much or the entire extent of the thoracoabdominal aorta. All patients classified as having PAU showed similar radiographic findings to the exemplary cases presented in Figures 7-10. All had contrast-filled outpouchings and an ulcerated crater seen on angiography, TEE, CT, or MR imaging. Associated intramural hematoma and pleural effusions were often identified as well. Severe aortic arteriosclerosis and calcification were seen uniformly. Patients diagnosed with IMH had imaging studies consistent with the representative examples of CT and TEE in IMH presented in Figures 11 and 12. There was absence of a flap traversing the aortic lumen, and imaging studies demonstrated the concentric orientation of the blood-containing space around the aortic lumen. These imaging studies were in distinct contrast with the findings in classic aortic dissection presented in Figures 4-6. The absence of an ulcer crater distinguishes these images of IMH from PAU with intramural hematoma.
646
COADY et a1
Figure 10. A, MR image (sagittal view) of a patient with a penetrating atherosclerotic ulcer. The penetrating atherosclerotic ulcer is diagnosed by visualization of a distinct ulcer crater in the absence of an intimal flap or false lumen. 6,MR image (axial view) of a patient with a penetrating atherosclerotic ulcer.
A number of different imaging modalities were employed to establish the diagnosis and to achieve follow-up in patients with PAU and IMH; more than one method was used for every patient (Tables 1 and 2). CT was the most commonly employed imaging modality. CT was used in 17 patients with PAU and 17 patients with IMH (94.4% of all patients). MR imaging studies were employed in 10 PAU and 4 IMH patients (38.8% of all patients). TEE was obtained in 15 PAU patients and 15 IMH patients (83.3% of all patients). Aortic
angiography was employed in 3 PAU patients and 6 patients with IMH (25.0% of all patients).
Demographic Considerations and Presenting Clinical Features
Tables 1 and 2 summarize the demographic and clinical features of the 19 PAU and the 17 IMH patients. Table 3 presents age and other
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
Figure 11. CT scan of a patient with an intramural hematoma. Note the presence of a concentric aortic wall hematoma and absence of an intimal dissection flap or penetrating ulceration.
Figure 12. TEE of a patient with an intramural hematoma. Note the absence of intimal disruption or penetrating atherosclerotic ulceration.
647
Moderate None Moderate None Mild None
F
75 69 69
56 69 64
14 15 16
17 18 19
None None Moderate
Mild None None None None Moderate None Moderate None
None None None Moderate None None Moderate Moderate None
Severe Moderate None None Moderate None Severe None None
None None
None Moderate
None Moderate None
Severe Moderate None
Moderate Moderate None Moderate Severe Moderate Moderate None None
Moderate Severe
Severe None
Renal Pulmonary Dysfunction Disease
-
Midscapular pain Mild left chest and midscapular pain Midscapular pain Left chest pain and hypotension Left chest pain Asymptomatic Midscapular pain Left chest pain Asymptomatic Asymptomatic Left chest pain Midscapular pain Anterior chest pain Midscapular pain Substemal pain Epigastric and chest pain Midscapular pain Midscapular pain Midscapular pain
Presenting Symptom
no no no
no no Yes
Yes Yes Yes no no Yes no no no
Yes no
Yes Yes
AAA
+ +
+ + + + ++ + ++ + + + + + +
+ + + + + + + + +
-
-
-
+ +
-
+ +
+ +
CT MRI TEE Angio
Management
Desc Desc Desc
Desc Desc Desc
Arch/Desc Asc Desc Desc Desc Desc Desc Desc Asc
Desc Desc
Surg Surg Surg
Surg Med Surg
Med Med Med Surg Med Med Med Surg Surg
Med Med
Desc Med Arch/Desc Med
Site
Rupture / Discharged Discharged Discharged
Rupture / Discharged Discharged Rupture / Discharged
Discharged Discharged Discharged Rupture / Discharged Discharged Discharged Discharged Discharged Rupture/Death
Rupture/Death Rupture/Death
Discharged Rupture/Death
Outcome
HTN = Hypertension; CT = computerized tomography; MRI = magnetic resonance imaging; TEE = transesophageal echocardiography; F = female; M = male; Desc = descending thoracic aorta; Asc = ascending aorta. Risk factors graded as per methods section. + Indicates study was performed for either diagnosis or follow-up. - Indicates study not performed. Med = medical management with antihypertensives and afterload reduction; Surg = patient underwent an operation. "Discharged implies that the patient was treated successfully and released from the hospital. "Rupture/Death" implies that the patient had an acute rupture and died in the hospital. "Rupture/Discharge" implies that the patient had an acute rupture prior to surgery, underwent resection of the penetrating ulcer, and was later released from the hospital.
F
F M
M F
Moderate Moderate Severe Moderate Severe Moderate Severe Severe Moderate
M M M F F F M F F
63 82 64 72 83 77 79 82 84
5 6 7 8 9 10 11 12 13
Mild None
Mild Severe
M M
76 85
3 4
Moderate None Severe Mild
Cardiac Disease
F M
Htn
77 79
Sex
1 2
Patient Age
Table 1. DEMOGRAPHIC AND CLINICAL CHARACTERISTICS OF PATIENTS WITH PENETRATING ATHEROSCLEROTIC ULCERS OF THE THORACIC AORTA
F F F
M MF
87 63
67 68 69 82 73 66
68 70
65 82
86
5
6
7 8 9 10 11 12
13 14
15 16
17
None
F
None
None Moderate
Severe None
None None Moderate Severe None None
Moderate Severe
None None Moderate None
Cardiac Disease
None
None None
Severe None
None None None Mild Mild None
Mild Severe
None None None None
None
None None
Mild Severe
Severe None Mild Moderate None Severe
None Mild
None Severe None None
Renal Pulmonary Dysfunction Disease
Midscapular pain Substemal pain Midscapular pain Chest pain and respiratory distress Intrascapular pain Anterior chest pain/ confusion Anterior chest pain Midscapular pain Midscapular pain Midscapular pain Anterior chest pain Substemal chest pain Lower back pain Substemal chest pain Midscapular pain Substernal chest pain Substemal chest pain
Presenting Symptom
n
n n
n n
o
o o
o o
yes n o n o yes no n o
yes n o
n o yes yes n o
AAA
+ +
+
+
+
+
CT
+
+ +
+ +
+
+ +
+
+
+
+ +
-
-
-
-
-
+
-
-
-
+
-
-
MRI
+
+
+ +
+ +
+
+
+ +
+
+
+
+
+ -
TEE
-
Asc
Surg
Desc Surg Asc Surg
-
-
Surg Med
Surg Med Med Med Med Surg
Med Med
Med Med Med Surg
Management
Desc Desc
Asc Desc Desc Desc Desc Asc
Desc Desc
Desc Desc Desc Asc
Site
-
-
+ + + -
-
-
+
-
-
+
Angio
Rupture / Death
Discharged Rupture/Discharged
Rupture / Discharged Discharged
Rupture / Discharged Discharged Discharged Discharged Discharged Death
Discharged Rupture/ Death
Discharged Discharged Discharged Rupture / Death
Outcome
HTN = Hypertension; CT = computerized tomography; MRl = magnetic resonance imaging; TEE = transesophageal echocardiography; F = female; M = male; Desc = descending thoracic aorta; Asc = ascending aorta. Risk factors graded as per methods section. + Indicates study was performed for either diagnosis or follow-up. - Indicates study not performed. Med = medical management with antihypertensives and afterload reduction; Surg = patient underwent an operation. "Discharged implies that the patient was treated successfully and released from the hospital. "Rupture/Death" implies that the patient had an acute rupture and died in the hospital. "Rupture/Discharge" implies that the patient had an acute rupture prior to surgery, underwent resection of the penetrating ulcer, and was later released from the hospital.
Mild Moderate
Moderate Moderate
Moderate Moderate None Moderate Mild Moderate
F M
M M
F F
M F F M
Moderate Moderate Severe Moderate
F
84 77 74 77
1 2 3 4
Moderate Severe
Htn
Sex
Age
Patient
Table 2. DEMOGRAPHIC AND CLINICAL CHARACTERISTICS OF PATIENTS WITH INTRAMURAL HEMATOMA OF THE THORACIC AORTA
650
COADY et a1
Table 3. COMPARISON OF TYPE A AND B AORTIC DISSECTIONS WITH INTRAMURAL HEMATOMAS AND PENETRATING ULCERS ~~
Type A Dissection
Patients Age Initial Aortic Size Prior AAA Rupture
80 (48 M, 32 F)
56.5 years ? 13.8 5.38 cm f 1.3 8 (10.0%) 6 (7.5%)
~
Type B Dissection
Penetrating Ulcer
Intramural Hematoma
98 (68 M, 30 F) 69.9 years* f 14.1 5.1 cm ? 1.3 30 (30.6%)$ 4 (4.1%)
19 (9 M, 10 F) 73.9 years* f 8.3 6.2 cmt ? 1.5 8 (42.1%)$ 8 (42.1%0)§
17 (7 M, 10 F) 74.0 years* f 7.8 5.5 cm -C 1.6 5 (29.4%) 6 (35.3%)§
Table reports means f standard deviation. Statistical analysis based on ANOVA using Bonferroni's method for multiple comparisons. = male; F = female. *The mean age for patients with type B dissection, penetrating ulcer, and intramural hematoma is significantly older than that of patients with type A dissection (P = 0.0001). tThe mean aortic diameter at initial presentation for patients with penetrating ulcer is significantly larger than for patients with type B dissection (P = 0.01). $The incidence of prior AAA for patients with penetrating ulcer is significantly greater than patients with type A dissection (P = 0.05). The incidence of prior AAA for patients with type B dissection is significantly greater than for patients with type A dissection M
(P =
0.01).
§The rupture rate for patients with penetrating ulcer is significantly higher than that of type A or B dissections ( P = 0.0001). The rupture rate for patients with intramural hematoma is significantly higher than either type A (P = 0.01) or type B dissection ( P = 0.001).
demographic and clinical data on the four categories of patients. Comorbidity
Accompanying medical problems were common in both PAU and IMH patients. For PAU, comorbidities included hypertension (18 patients, 94.7%), chronic obstructive pulmonary disease (12 patients, 63.1%), cardiac disease (8 patients, 42.1%), chronic renal insufficiency (5 patients, 26.3%), and diabetes mellitus (3 patients, 15.7%) (see Table 1).For IMH patients, comorbidities included hypertension (15 patients, 88.2%), chronic obstructive pulmonary disease (8 patients, 47.1%), cardiac disease (7 patients, 41.2%), and chronic renal insufficiency (5 patients, 29.4%) (Table 2). Pain Presentation
PAU and IMH patients presented almost uniformly with excruciating chest or back pain, similar to that of aortic dissection. As with classic dissection, anterior chest pain correlated with ascending aortic events and back pain with descending aortic events (see Tables 1 and 2). Ischemic Complications
No PAU or IMH patients manifested the branch vessel occlusion or run-off organ is-
chemia seen commonly with classic dissection.
Age Profiles
A distinct difference in the age profile was seen between classic dissection patients and those with PAU or IMH (see Table 3). Type B dissectors, PAU patients, and IMH patients were significantly older (means, 69.9, 73.9, and 74.0 years, respectively) than type A dissectors (mean 56.5 years) ( P = 0.0001). Although not statistically significant, patients with PAU and IMH tend to be older than those with type B dissection. Furthermore, PAU and IMH occurred within a narrow age range of elderly patients (63-85 and 63-87 years, respectively) (Fig. 13).
Aortic Sizes
All four aortic pathologies occurred in enlarged aortas (mean 5.3 cm). There were differences in aortic size at the time of occurrence of the various pathologic lesions. Patients with type A dissection, type B dissection, and IMH had similar initial aortic diameters (means 5.4, 5.1, and 5.5 cm, respectively). Patients with PAU had the largest aortas at presentation, measuring 6.2 cm (when compared to patients with type B dissection, P = 0.01) (see Table 3).
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
25
651
Type A Dissection
1
2ol
3 20-
8 &
. c) I
LI
0
z E 2
155 1
1
10-
5-
o', 20
I
30
I
I
I
40
B
I
I
50 Ages
I
60
I
I
70
I
I
I
80
Figure 13. A, Comparison of age distributions for type A dissections, penetrating atherosclerotic ulcer, and intramural hematoma. 6,Comparison of age distributions for type B dissections, penetrating atherosclerotic ulcer, and intramural hematoma.
Concomitant Abdominal Aortic Aneurysm
Dramatic differences were found in the incidence of concomitant abdominal aortic aneurysms (AAA). Patients with type B dissections, PAU, and IMH had much higher incidence of concomitant AAA (30.6%, P = 0.01,42.1%, P = 0.05, and 29.4%, respectively) when compared to type A dissection (10Y0). Patients with PAU had the highest incidence of accompanying AAA (42.1%) (see Table 3). Location
PAU was almost entirely a descending aortic phenomenon, with 17 of 19 (89.5%) occurring in the descending aorta. In 2 of these 17 patients, the hematoma extended subadventitially proximal to the ligamentum arteriosum into the aortic arch. The remaining 2 patients (10.5%) had ulcers located in the ascending aorta. IMH was also largely a descending aortic phenomenon, with 12 of 17
(70.6%) hematomas located in the descending aorta. Hospital Course
Figure 13 presents the incidence of rupture for all four lesions: type A dissection, type B dissection, PAU, and IMH. PAU were found to rupture much more commonly than either classic type of dissection (P = 0.0001). Similarly, IMH were found to rupture much more commonly than type A (P = 0.01) or type B dissection (P = 0.001) (Figure 14, Table 3). The incidence of acute rupture for PAU is 42.1% (8 of 19 patients) as opposed to 35.3% for IMH (6 of 17 patients), 7.5% (6 of 80 patients) for type A dissections, and 4.1% (4 of 98 patients) for type B dissections (P = 0.0001). All ruptures occurred with initial presentation or during initial hospitalization. With only one exception, all patients with PAU and IMH in the ascending aorta ruptured (see Tables 1 and 2). No ruptures occurred in late follow-up.
652
COADY et a1
(8/19)
(6/17)
(6/80)
Type A
Type B
Penetrating Ulcer
Intramural Hematoma
Figure 14. Rupture rates for type A dissections, type B dissections, penetrating atherosclerotic ulcer, and intramural hematoma.
Of the PAU patients, 8 of 19 (42.1%) ruptured prior to surgical treatment. Three individuals (Patients 2, 3, and 4 in Table 1) were thought initially to have a type B aortic dissection and died while being treated with standard medical therapy (P-blockade and afterload reduction) in the intensive care unit. Confirmatory autopsies were obtained in these three patients. The other 5 patients who ruptured and 3 additional patients without evidence of rupture (Patients 8, 12, 13, 14, 16, 17, 18, 19 in Table 1) were taken urgently or emergently to the operating room, with 7 (87.5%)surviving to discharge. The remaining 8 medically treated patients were discharged from the hospital following medical therapy. Of the IMH patients, 6 of 17 (Patients 4, 6, 7, 13, 16, and 17) (35.3%) ruptured prior to surgical treatment. An additional two IMH patients demonstrated expansion of their hematoma during the initial hospitalization (Pa-
tients 4 and 6, Table 2). Therefore, 8 of 17 IMH patients (47.1%) behaved malignantly during the initial presentation. Seven of these 8 patients underwent operations urgently or emergently, with 4 (50%) surviving to discharge; the eighth patient died of rupture prior to surgery. Pathology
In PAU, the gross pathology was confirmatory of surgical and radiographic findings. The aorta appeared densely atherosclerotic. The microscopic findings demonstrated intima1 degeneration and replacement with cholesterol clefts, which burrowed thorough the aortic media (Color Fig. 15). Depending upon the depth of penetration in specific patients, blood was found at varying levels within the media or beyond. The ulcer at times precipitated a localized intramedial dissection asso-
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
ciated with a variable amount of hematoma within the aortic wall. In IMH, gross findings in the operating room included a very superficial location of the hematoma in the aortic wall. In the operating room, in contradistinction to classic dissection, after the external adventitial layer was opened, the inner layer remained convex and appropriate in shape, not falling away from the aortic wall in concave collapse as seen in typical dissection. The IMH pathology uniformly demonstrated a significant degeneration of the aortic media, with hematoma located just cells away from the adventitia (Color Fig. 16). This finding is in contradistinction to our usual pathology findings in classic dissection, where the plane of hematoma is routinely found not more than two-thirds of the way through the media (see Figure 1).
653
radiologic follow-up, 3 individuals (Patients 8, 11, 14, Table 2) showed involution of their hematoma, 1 progressed to typical dissection with an intimal flap (Patient 3, Table 2) and 3 demonstrated no change in their hematomas over time (Patients 5, 9, and 10). All surgically treated patients surviving to discharge are well with no further aortic complications. Figure 17 displays the survival curves for all 19 PAU and all 17 IMH patients, compared to survival for the more common type A and type B dissectors. IMH patients had the lowest 1-year survival (65.8%) when compared to type A (88.9%) and type B (90.7%) dissectors ( P = 0.03). Although not achieving statistical significance, the 1-year survival for PAU patients (76.4%) falls short of the survival for both type A and type B dissectors. SUMMARY
Late Follow-up
Of the 4 surgical and 8 medically treated PAU patients surviving to discharge, all remain alive at mean follow-up of 31.5 months (range 3 to 84 months). On serial imaging, one individual (Patient 15, Table 1) has shown reduction in the size of his intramural hematoma. Of the 4 surgical and 9 medical IMH patients surviving to discharge, 2 died of unrelated causes (Patients 1 and 2, Table 2), and the others remain alive at mean follow-up of 26.4 months (range 1to 52 months). On serial
This article confirms the existence of two variants of acute aortic pathology, the penetrating atherosclerotic ulcer (PAU) and the intramural hematoma (IMH), which are radiologically distinct from classic aortic dissection. Table 4 reviews the characteristicsdistinguishing PAU from classic aortic dissection and IMH. We took as a matter of definition that classic aortic dissection involves a flap which traverses the aortic lumen. We defined PAU and IMH as nonflap lesions, with PAU demonstrating a crater extending from the
* Intramural Hematoma -
h
m
0.6 -
$
0.4
.-5
-
v)
0
3
6
9
12
Months
Figure 17. Kaplan-Meier 1-year survival for patients with penetrating atherosclerotic ulcer, intramural hematoma, type A, and type B aortic dissections. Asterisk indicates that P = 0.03.
654
COADY et a1
Table 4. CHARACTERISTICS DISTINGUISHING PENETRATING ULCER AND INTRAMURAL HEMATOMA FROM AORTIC DISSECTION Characteristic
Penetrating Ulcer
Intramural Hematoma
No intimal flap Intramural hematoma Localized ulceration penetrating internal elastic lamina Elderly hypertensive Hypertensive
No intimal flap No penetrating ulceration Hematoma within aortic wall
Intimal flap with contrast filling false lumen
Elderly Hypertensive
Same
Signs
Severe chest or midscapular pain Absent No branch vessel occlusion
Younger Hypertensive Occasionally bicuspid aortic valve Marfan’s syndrome Same
Absent No branch vessel occlusion
Location
Largely descending
Largely descending
Extent of lesion Degree of atherosclerosis Aortic size Concomitant AAA
Focal Always severe Largest (-6.2 cm) Common
Focal Variable Larger (-5.5 cm) Common
Diagnostic features
Typical patient
Symptoms
Aortic Dissection
Aortic insufficiency (type A) Pulse inequality Neurologic deficits compromise of blood flow to branch vessels Even distribution of ascending and descending Usually extensive Variable (often minimal) Smaller (-5.2 cm) Occasional (except type B, common)
AAA = abnormal aortic aneurysm.
aortic lumen into the space surrounding the aortic lumen. This categorization can be summarized with the expression, ”no flap, no dissection.” With these definitions made, re-review of the imaging studies for the present report identified 36 such lesions out of 214 cases originally read as aortic dissection. Therefore, these variant lesions accounted for over 1 out of 8 acute aortic pathologies. Besides confirming the existence of the conditions, PAU and IMH, as distinct radiographic lesions, this series strongly suggests that these two conditions constitute distinct clinical entities as well. Table 4 summarizes the clinical patterns of these two entities as apparent from the present study, and contrasts them with classic aortic dissections. In particular, the following observations, some of which are consonant findings in smaller series,+ can be made regarding the typical patient profiles of PAU and IMH from the present study: *References 1, 4, 11, 14, 15, 19, 23, 28
The patients with PAU and IMH are distinctly older than those with type A aortic dissection (74.0 and 73.9 versus 56.5 years, P = 0.0001). Although not statistically significant, PAU and IMH patients tend to be older than patients with type B aortic dissections as well. For PAU and IMH, unlike aortic dissection, the concentration in the elderly is manifested in a very small standard deviation of the mean age (see Fig. 13); these two entities, PAU and IMH, are essentially diseases of the seventh, eighth, and ninth decades of life. Patients with PAU and IMH are almost invariably hypertensive (about 94% of cases). The pain of PAU and IMH mimics that of classic aortic dissection, with anterior symptoms in the ascending aortic lesions and intrascapular or back pain with descending aortic lesions. Unlike classic dissection, PAU and IMH do not produce branch vessel compro-
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
mise or occlusion and do not result in ischemic manifestations in the extremities or visceral organs. PAU and IMH are more focal lesions than classic aortic dissection, which frequently propagates for much or the entire extent of the thoracoabdominal aorta. PAU is uniformly associated with severe aortic arteriosclerosis and calcification, whereas classic dissection often occurs in aortas with minimal arteriosclerosis and calcification. PAU and IMH tend to occur in even larger aortas than classic aortic dissection (6.2 and 5.5 versus 5.2 cm, P = 0.01). PAU and IMH are strongly associated with AAA, which is seen concomitantly in 42.1% of PAU patients and 29.4% of IMH patients. PAU and IMH are largely diseases of the descending aorta (90% for PAU and 71% for IMH). Although our pathology data is limited, we do feel that an inherent difference in the histologic intramural level of the hematoma may underlie the pathophysiologic process that determines which patient develops a typical dissection and which develops an intramural hematoma. In particular, we feel that the level of blood collection is more superficial, closer to the adventitia, in IMH than in typical aortic dissection. This may explain why the inner layer does not prolapse into the aorta on imaging studies or when the aorta is opened in the operating room. This more superficial location would also explain the high rupture rates as compared to classic aortic dissection (Fig. 14, Table 3). We did find PAU and IMH to behave much more malignantly than typical descending aortic dissection. As seen in Figure 6, the rupture rate is much higher than for aortic dissection. Documented rupture occurred in 35.3% of IMH patients and in 42.1% of PAU patients. All but one PAU or IMH present in the ascending aorta ruptured. These ruptures all occurred during the initial hospitalization. These rupture rates are much higher than the 7.5% seen for ascending dissection and the 4.1% seen for descending dissection (P = 0.0001). (Of course, we do not ordinarily ob-
655
serve ascending dissections without surgery long enough to allow them to manifest their known poor prognosis). Therefore, PAU and IMH are extremely malignant in their early behavior. For this reason, it is essential to differentiate PAU and IMH from typical aortic dissection. Once patients with PAU and IMH survive the initial hospitalization, they do well, without late symptom recurrence or rupture in follow-up (mean 28 months in our study). What therapeutic correlates can we draw from this experience? First, detection is important. IMH and PAU are life-threatening conditions, which may not be detected on presentation unless the clinician is aware of the distinguishing radiologic features. These are enumerated in the previous paragraphs. Second, the following suggestions are made regarding management: We recommend early operative intervention for all PAU and IMH affecting the ascending aorta. Ascending lesions, although uncommon, in our experience uniformly ruptured or progressed and were often lethal. Therefore, these patients may be treated as those with classic ascending aortic dissections. This posture is supported by other series.'" 17, 28 Patients with PAU and IMH lesions in the descending aorta can be observed initially and treated with "antiimpulse" therapy, including P-blockade and afterload reduction. These lesions are more serious than classic descending aortic dissection, and a low threshold for surgical intervention must be maintained. If radiographic findings are ominous (severely bulging hematoma, extensive subadventitial spread, extra-adventitial blood, bloody pleural effusion, or deeply penetrating ulcer), surgery should be performed preemptively. In the absence of ominous initial radiographic findings, repeat imaging should be carried out within 3 to 5 days. In case of any progression, in volume or extent of the hematoma or overall size of the aorta, surgery should be performed preemptively to prevent rupture. Likewise, if symptoms are not controlled or recur on medical treatment, surgery
656
COADY et a1
should be performed. This posture is supported by most other series.14,15, 23, 28 If patients tolerate early medical management without clinical deterioration, our experience suggests that they may continue to be followed conservatively. In long-term follow-up, general size criteria for surgical intervention on the thoracic aorta can be a ~ p l i e d . ~ Kouchoukos and Dougenis recently reviewed strategies for evaluation and treatment of aortic pathologies.12Our findings and clinical recommendations are consonant with those reported in smaller series of variant aortic pathologies examined in that review. As IMH and PAU primarily affect older individuals with significant comorbidity, however, our above recommendations must be tempered by consideration of the patient’s overall condition and general outlook. Some of these patients may not be appropriate candidates for the aggressive surgical strategy outlined. We are just beginning to understand the identity and behavior of these lesions, and our management suggestions are provisional recommendations regarding the treatment of penetrating ulcer and intramural hematoma. Continued experience will be required to assess more completely the natural history of these variants of aortic dissection and to achieve more definitive treatment recommendations. Furthermore, although we have described these variant lesions as entities distinct from aortic dissection, it is likely that intermediate lesions exist, and that ulcer, hematoma, and dissection lie on a continuum of disease processes. References 1. Alfonsa F, Goicolea J, Aragoncillo P, et al: Diagnosis of aortic intramural hematoma by intravascular ultrasound imaging. Am J Cardiology 76:735-738, 1995 2. Coady MA, Rizzo JA, Hammond GL, et al: Penetrating ulcer of the thoracic aorta: What is it? How do we recognize it? How do we manage it? J Vasc Surg 271006-1016, 1998 3. Coady MA, Rizzo JA, Hammond GL, et al: What is the appropriate size criterion for resection of thoracic aortic aneurysms? J Thorac Cardiovasc Surg 113:47& 491, 1997
4. Cooke JP, Kazmier FJ, Orszulak TA: The penetrating aortic ulcer: Pathologic manifestations, diagnosis, and management. Mayo Clin Proc 63:718-725, 1988 5. Crawford ES, Svensson LG, Coselli JS, et al: Aortic dissection and dissecting aortic aneurysms. Ann Surg 208(3):254-273, 1988 6. Gore I: Pathogenesis of dissecting aneurysms of the aorta. Arch Pathol Lab Med 53:142-153, 1952 7. Gore I, Seiwert VJ: Dissecting aneurysm of the aorta. Arch Pathol 53:121, 1952 8. Hirst AE, Gore I: Is cystic medionecrosis the cause of dissecting aortic aneurysm? (editorial). Circulation 53(6):915-916, 1976 9. Hirst AE, Gore I: The etiology and pathology of aortic dissection. In Doroghazi RM, Slater EE (eds): Aortic Dissection. New York, McGraw-Hill, 1983 10. Hirst AE Jr, Johns VJ Jr, Klime SW Sr: Dissecting aneurysm of the aorta: A review of 505 cases. Medicine (Baltimore) 37217-279, 1958 11. Kazerooni EA, Bree RL, Williams DM: Penetrating atherosclerotic ulcers of the descending thoracic aorta: Evaluation with CT and distinction from aortic dissection. Radiology 183:759-765, 1992 12. Kouchoukos NT, Dougenis D. Surgery of the thoracic aorta. NEJM 26:1876-1888, 1997 13. Larson EW, Edwards WD: Risk factors for aortic dissection: a necropsy study of 161 cases. Am J Cardiol 53(6):849-855, 1984 14. Mozsowitz HD, Lampert C, Jacobs LE, et al: Penetrating atherosclerotic aortic ulcers. Am Heart J 128:1210-1217, 1994 15. Muluk SC, Kaufman JA, Torchiana DF, et al: Diagnosis and treatment of thoracic aortic intramural hematoma. J Vasc Surg 24:1022-1029, 1996 16. Murray CA, Edwards JE: Spontaneous laceration of the ascending aorta. Circulation 472348-858, 1973 17. Murray JG, Manisali M, Flamm SD, et al: Intramural hematoma of the thoracic aorta: MR image findings and their prognostic implications. Radiology 204:349355, 1997 18. Nakashima Y, Kurozami T, Sueishi K, et al: Dissecting aneurysm: A clinicopathologic and histopathologic study of 111 autopsied cases. Hum Pathol 21L:291296, 1990 19. Neinaber CA, von Kodolitsh Y, Peterson B, et al: Intramural hemorrhage of the aorta: Diagnostic and therapeutic implications. Circulation 92:1465-1472, 1995 20. Patterson DL, Brennan S, Cartwright T, et al: Traumatic rupture of an aortic ulcerative atherosclerotic plaque producing aortic dissection: A complication of interscapular back blows used to dislodge objects from the esophagus. Clinics in Cardiology 16:741744, 1993 21. Roberts WC: Aortic dissection: Anatomy, consequences, and causes. Am Heart J 101:195-214, 1981 22. Roberts WC, Honig HS: The spectrum of cardiovascular disease in the Marfan syndrome: A clinicomorphologic study of 18 necropsy patients and comparison to 151 previously reported necropsy patients. Am Heart J 104:115, 1982 23. Schappert T, Sadony V, Schoen F, et al: Diagnosis and therapeutic consequences of intramural aortic hematoma. J Card Surg 9:50&515, 1994 24. Schnitker MA, Bayer CA: Dissecting aneurysm of aorta in young individuals, particularly in associated with pregnancy. Ann Intern Med 20:486, 1944 25. Shachter N, Perloff JK, Mulder DG: Aortic dissection in Noonan’s syndrome (46 XY Turner). Am J Cardiol 54(3):464-465, 1984
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS 26. Shennan T Dissecting aneurysms. Medical Research Council, Special Report Series, No. 193, 1934 27. Slater DN, Grundman MJ, Mitchell L: Turner’s syndrome associated with bicuspid aortic stenosis and dissecting aortic aneurysms. Postgrad Med J 58(681):436-438, 1982 28. Stanson AW, Kazmier FJ, Hollier LH, et al: Penetrating atherosclerotic ulcers of the thoracic aorta: Natural history and clinicopathological correlations. Ann Vasc Surg 1:15-23, 1986 29. Stolle CA, Pyeritz RE, Myers JC, et al: Synthesis of an altered type I11 procollagen in a patient with type IV Ehlers-Danlos syndrome: A structural change in the alpha 1 (111) chain which makes the protein more susceptible to proteinases. J Biol Chem 60:1937, 1985
657
30. Svensson LG, Crawford ES: Occlusive disease of the aorta. In Svensson LG, Crawford ES (eds): Cardiovascular and Vascular Disease of the Aorta. Philadelphia, WB Saunders Company, 1997 31. Wheat MW. Pathogenesis of aortic dissection. In Doroghazi RM, Slater EE (eds): Aortic Dissection. New York, McGraw-Hill, 1983, p 55 32. Wolinsky H: Response of the rat aortic media to hypertension: Morphological and chemical studies. Circ Res 26(4):507-522, 1970 33. Yamada T, Takamiya M, Naito H, et al: Diagnosis of aortic dissection without intimal rupture by x-ray computed tomography. Nippon Acta Radio1 45:699710, 1985
Address reprint requests to John A. Elefteriades, MD Section of Cardiothoracic Surgery Yale University School of Medicine 333 Cedar Street New Haven, CT 06510
~
~
07334651199 $8.00
DISEASES OF THE AORTA
+ .OO
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS Penetrating Atherosclerotic Ulcers and Intramural Hematomas Michael A. Coady, MD, MPH, John A. Rizzo, PhD, and John A. Elefteriades, MD
Penetrating atherosclerotic ulcer (PAU) and intramural hematoma (IMH) of the aorta were virtually unknown in the prior era of aortic imaging by aortography. These disorders were not defined as radiographically dis33 In the current tinct until the mid-1980~.~~, era of three-dimensional, high resolution imaging of the aorta by computerized tomography (CT), magnetic resonance (MR) imaging, and transesophageal echocardiography (TEE), these two disorders have become increasingly recognized. The distinctions from typical aortic dissection have not always been clear in terms of diagnosis, clinical characteristics, and treatment. Most reports of PAU and IMH in the literature have concentrated on identifying these disorders as entities distinct from classic aortic dissection by describing small series of these patients.* There is growing concern that the anatomic pathology and clinical behavior of PAU and IMH may differ substantially from those of typical aortic dissection, and that clinical management may need to be spe*References 1, 4, ll, 14, 15, 19, 23, 28
cifically tailored. In particular, it is becoming more apparent that the paucity of information and relative rarity of diagnosis are due to the fact that cases of PAU and IMH are often misdiagnosed and managed as "aortic dissections." The infrequence, misdiagnosis by initial radiologic inspection, and lack of clinical experience with PAU and IMH has not permitted definition of optimal therapy for these two variants of aortic disease. The present review draws on our experience at the Yale Center for Thoracic Aortic Disease in order to clarify the disorders of PAU and IMH of the aorta. This article expands on our initial report2 and reviews the clinical features, radiographic appearance, pathologic findings, and natural history of individuals with PAU and IMH. Appropriate correlates regarding the optimal treatment of these individuals are also discussed. CLASSIC AORTIC DISSECTION
A classic aortic dissection begins with a laceration of the aortic intima and inner layer
From the Department of Surgery, Section of Cardiothoracic Surgery (MAC, JAE), and the Department of Epidemiology and Public Health (JAR), Yale University School of Medicine, New Haven, Connecticut
CARDIOLOGY CLINICS OF NORTH AMERICA VOLUME 17 * NUMBER 4 NOVEMBER 1999
637
638
COADY et a1
of the aortic media, which allows blood to course freely along a false lumen in the outer third of the media (Color Fig. 1). An artist’s rendition of a typical aortic dissection is presented in Fig. 2, and PAU and IMH are illustrated for later comparison. As is easily appreciated, a classic aortic dissection involves a dissection flap that traverses the aortic lumen, dividing the aorta into true and false lumina. The artist’s rendition corresponds well to the operative findings familiar to the aortic surgeon. The flap traverses the aortic lumen obliquely in cross-sectional views; it is not oriented circurnferentially. This standard appearance of classic aortic dissection will later be contrasted with imaging findings in PAU and IMH in our patients. Understanding the pathogenesis of aortic dissection requires consideration of the inciting event causing the propagation of blood within the aortic media. Although various risk factors predisposing the aorta to dissection have been well-described, the precise insult leading to laceration of the intima and media remains unclear. Historically, the primary etiologic event leading to aortic dissection has been extremely controversial. Cystic medial necrosis associated with Marfan’s disease and other connective tissue disorders was once believed to contribute to aortic medial degeneration, leading to aortic dissection. Larson and Edw a r d ~ , ’however, ~ demonstrated that only a few of their 161 patients with known aortic dissection exhibited medial degeneration. They found that 158 of these autopsies had intimal aortic tears, supporting the theory initially proposed by Murray and Edwards,I6 that the intimal tear is the primary event,
allowing the blood to spread through the outer two-thirds of the aortic media. Cystic medial necrosis is no longer regarded as the common structural disorder underlying aortic dissection. Gore6in 1952 suggested that the etiology is related to spontaneous rupture of the aortic vasa vasorum leading to subsequent intimal tear and aortic dissection. He accumulated 29 cases of aortic dissection without intimal tear, describing a process in which IMH precedes intimal rupture because hemorrhage of the vasa vasorum weakens the media, allowing blood to leak into the media from the aortic lumen owing to arterial pressure. Degenerative changes within the media itself can affect its elastic constituents. Loss of these elements reduces the resistance of the aortic wall to hemodynamic stress, leading to aneurysmal dilation of the vessel with subsequent dissection. Our group at the Yale Center for Thoracic Aortic Disease has recently confirmed the increasing incidence of aortic dissection with greater aortic diameters (Figure 3).3In the dilated aorta, the stress leading to aortic dissection is composed of a radial and a longitudinal vector. As the aorta becomes increasingly more aneurysmal, the longitudinal vector becomes the more significant force. This explains the typical occurrence of a transverse tear along the course of the aortic wall. Inheritable disorders of elastic tissues also predispose to the development of aortic dissections, including Marfan,22Turner,27Nooand Ehlers-Danlos S y n d r ~ m e sOf .~~ these Marfan Syndrome is the most common and best understood. Other known risk factors for the development of aortic dissection
Figure 1. Classic aortic dissection. Hematoxylin-eosin stain of aortic wall (original magnification x 100). Note the medial disruption with blood seen coursing through the outer two-thirds of the aortic media. Figure 15. Aortic wall (hematoxylin-eosin, original magnification x 100) of a patient with rupture of a penetrating atherosclerotic ulcer. The aortic intima has been replaced with cholesterol clefts that have burrowed though the aortic media. Blood is seen on the outside of the media. There is no remaining adventitial layer. Figure 16. Aortic wall (hematoxylin-eosin, original magnification x 100) of a patient with an intramural hematoma. Note the lack of intimal disruption and medial disruption from blood within this layer.
639
640
COADY et a1
Figure 2. Aortic dissection, intramural hematoma, and penetrating atherosclerotic ulcer. Note the distinct intimal flap separating the true and false lumens in aortic dissection. An intramural hematoma, possibly caused by rupture of the vasa vasorum, leads to a concentric hematoma within the media, and has no intimal tear. Penetrating atherosclerotic ulcers burrow deeply through the intima into the aortic media, precipitating a localized intramedial dissection associated with a variable amount of hematoma.
include pregnancyz4and inflammatory diseases of the media.5 Significant derangement and loss of elastic tissue are common in younger patients with aortic dissection, especially those with congenital defects; however this correlation in older patients has not held welL9Some investigators believe that a loss of smooth muscle cells, often appearing in a laminar pattern, is the predominant factor leading to aortic dissection in patients older than 40 years.7 Because degenerative changes do not fully explain the occurrence of the dissection, various mechanical forces have also been considered. These include flexional forces of the ves-
sel at fixed sites, the radial impact of the pressure pulse, and the shear-stress of the blood. The heart, ascending aorta, and aortic arch are relatively mobile, whereas the descending aorta is suspended to the spinal column. During the cardiac cycle, the heart and aorta produce rhythmic movements, allowing all but fixed segments to move.31These fixed points are exposed to the most significant flexional forces. Classic type A and B aortic dissections may produce an intimal tear at the areas of greatest hydraulic stress: the right lateral wall of the ascending aorta, or the descending thoracic aorta in proximity to the ligamentum arteriosum. These are the most
(6/7 1) 8.5% (4/47) 8.5%
( 12/42) 28.6%
(1/14)
7.1%
n -
5
26
Thoracic Aortic Size (a) Figure 3. Increasing incidence of aortic dissection with larger aortic sizes. (From Coady MA, Rizzo JA, Hammond GL, et al: What is the appropriate size criteria for resection of thoracic aortic aneurysms? J Thorac Cardiovasc Surg 113:476491, 1997.
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
641
Figure 4. Computerized tomography (CT) of a patient with a classic type B aortic dissection. Note the intimal flap and presence of both true and false lumina.
common sites of aortic dissection, emphasizing the significant impact that flexional forces have on the aortic wall. Hypertension and increased aortic blood pressure alters the slope of the pulse pressure wave, double-pole double-throw (dp / dt), and has been shown to add a mechanical strain on the aortic wall leading to deposition of additional elastic lamellae.8,32 There is also an increased propensity for delamination of the aortic wall, because of the shearing forces exerting a longitudinal stress along the aortic wall. At one time it was also thought that atherosclerosis was the cause of acute aortic dissections. Among the earliest believers in this theory were Virchow (1851), Girode (1887), Ewald (1890), Rolleston (1893), and von Moller (1906).26In a classic paper entitled ”Dissecting Aneurysms,” Shennan in 1934 began to recognize that although nodular atheromas were present in a large number aortic dissections and ruptures, only a few displayed a relationship of the atheroma to the exact location of primary rupture or dissectionZ6Only 6 of 218 dissections in his study demonstrated evidence of a dissection at the base of an atheromatous ulcer. There still remains some controversy as to whether atherosclerosis causes aortic dissection.’” Most authorities believe it does not.
Those who believe that atherosclerosis causes aortic dissection do not reason that aortic dissections are more frequently found in the ascending aorta, where atherosclerosis is less common. Additionally, dissection in an area of gross atherosclerosis is usually limited, because of neighboring fibrosis and calcification. The imaging studies useful in diagnosis classic aortic dissections demonstrate two or more channels separated by an intimal flap on CT scans (Fig. 4), ECHO (Fig. 5), and angiography (Fig. 6). These images are presented to permit later comparison with findings in PAU and IMH. PENETRATING ATHEROSCLEROTIC ULCERS
In 1934, Shermanz6was the first to describe penetrating atheromatous ulcers of the thoracic aorta. These atheromatous plaques, characterized further by Stanson et aIz8in 1986, ulcerate and disrupt the internal elastic lamina, burrowing deeply through the intima into the aortic media.4 The plaque may precipitate a localized intramedial dissection associated with a variable amount of hematoma within the aortic wall, may break through into the adventitia forming a pseudoaneurysm,
642
COADY et a1
Figure 5. A, Transesophageal echocardiogram (TEE) showing a transverse plane at the level of the diaphragm showing a dissection flap. B, TEE showing a long axis plane at the level of the aortic valve, showing an extensive aortic dissection anteriorly and posteriorly. LA = left atrium.
or may rupture completeIy into the right or left hemithorax. Ulceration of an aortic atheroma occurs in patients with advanced atherosclerosis; however, this process is usually asymptomatic, confined to the intimal layer, and not associated with an intramural hematoma. When an atherosclerotic plaque penetrates into the media, the media is exposed to pulsatile atrial flow, causing hemorrhage into the wall without an intimal flap. This PAU involves a localized dissection, and the intramural hematoma is prevented from extension by surrounding transmural inflammation and relative fusion of the aortic wall layers.21The adventitial layer is frequently dissected from the media in patients who have penetration through the media.30The adventitia may also rupture, and
only the surrounding mediastinal tissues then contain the hematoma. Patients with PAU, therefore, have a localized dissection which is limited by areas of severe calcification associated with locally advanced atherosclerotic disease, therefore representing a different aortopathy than that of a classic aortic dissection. The site of entry into the dissection in patients with penetrating ulcers is the ulcer itself. On postmortem examination, there may be several atherosclerotic ulcers, often extending into the adventitia. The clinical presentation of patients with penetrating aortic ulceration is similar to that of classic aortic dissection; however, penetrating ulceration represents a unique pathologic event that may have distinct prognostic and therapeutic implication^.'^ Since preliminary reports4,14, ** indicate that penetrating ulcer is distinct from aortic dissection, our group recently sought to isolate true cases and study the natural history of these atheromatous ulcers. Because penetrating ulcer is much less common than classic aortic dissection, it may be often unrecognized, and therefore misclassified as an aortic dissection on presentation. Penetrating ulcers present similarly to typical aortic dissection, with chest or back pain of excruciating intensity. Absent, however, are findings of branch vessel occlusion or run-off organ ischemia, as PAU does not generally impair luminal blood flow. On imaging by CT, MR, and TEE, the cardinal findings for PAU include an ukev crater similar to the barium appearance of a duodenal ulcer. In the artist’s rendition (see Fig. 1) of PAU and IMH corresponds closely with the images engendered by their verbal designations. PAU is shown as an ulcer crater with varying degrees of surrounding hemorrhage. PAU does not have a flap traversing the aortic contour and defining a true and false lumen. These renditions corresponds well to current understanding of the pathophysiology of these disorders. The diagnosis of PAU is made on CT scan and angiography by demonstration of a contrast-filled outpouching in the aorta in the absence of a dissection flap or a false lumen, and often in the presence of extensive aortic calcification (Fig. 7 and 8).11TEE demonstrates a craterlike ulcer with jagged edges, usually
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
Figure 6. Angiography illustrating a classic aortic dissection. Note the dissection flap indicated by the arrows.
Figure 7. CT scan of a patient with a penetrating atherosclerotic ulcer at the level of the mid-descending thoracic aorta. Clearly demonstrated is a contrast filled ulcer. No intimal flap is seen.
643
644
COADY et a1
Figure 8. Angiography demonstrating a penetrating atherosclerotic ulcer.
in the presence of extensive aortic atheroma; there is absence of an intimal flap or a false lumen (Fig. 9). On MR imaging scans (Fig. lo), PAU are diagnosed by the visualization of a distinct ulcer crater in the absence of an intimal flap or false lumen. In order to confirm the existence of penetrating ulcers, a recent retrospective review of every patient admitted with a preliminary diagnosis of aortic dissection at Yale-New Haven Hospital from 1985 to 1998 was conducted.2Of 212 patients initially diagnosed as having an aortic dissection, 19 individuals (9%) (9 males and 10 females) were found to have PAU on re-review of CT, MR imaging, angiography, ECHO imaging studies, intraoperative findings, or pathology results. Demographic information, complications, age profiles, aortic sizes, location, hospital course, and follow-up were compared with classic dissections and intramural hematomas. INTRAMURAL HEMATOMAS (IMH) Although classical aortic dissections begin with an intimal tear which allows blood to
course rapidly along a plane in the outer third of an intrinsically diseased media, an IMH is thought to begin with rupture of the vasa vasorum of the aortic wall. This results in a circumferentially oriented blood-containing space seen on three-dimensional imaging studies (see Fig. 1). Note in Figure 1 that IMH is shown as a blood-filled circumferential space without ulcer or other intimal break or discontinuity. Krukenberg first described IMH in 1920 as a ”dissection without intimal tear.”19 This variation of aortic dissection was considered a distinct entity at postmortem examination. On histologic analysis, a hematoma is visualized disrupting the medial layer of the aorta. In 1985, Yamada et al,33 proposed the term ”aortic dissection without intimal rupture” to describe intramural hematomas in isolation of intimal disruption or penetrating ulceration. In their series, 6 of the 14 patients (43%) had involvement of the ascending aorta, and the remaining 8 patients (57%) had involvement the descending aorta. All patients were medically managed with serial imaging studies; 3 patients (21%) died within 1 month of diagnosis (2 ascending and 1 descending) and another died following a graft replacement. Nine of the 10 survivors showed complete resolution of the aortic intramural hematoma within 1 year. IMH may occur spontaneously in hypertensive patients, or following blunt chest trauma.’ It has been speculated that aortic intramural hematoma may also originate from ulceration in an atherosclerotic aorta; therefore, the cause may be m~ltifactorial.’~ These individuals may have no evidence of an atheromatous ulcer on imaging studies, however, and an intimal tear is not visualized. Because there is no intimal discontinuity, the space does not communicate directly with the aortic lumen and, unlike the false lumen of classic aortic dissection, does not generally show enhancement with contrast administration on CT scanning (Fig. 11) and angiography.19 On TEE examinations, IMH displays an aortic wall hematoma in the absence of an intimal dissection flap or penetrating ulceration (Fig. 12).
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
645
Figure 9. TEE demonstrating a penetrating atherosclerotic ulcer.
THE YALE EXPERIENCE: DISTINGUISHING PAU AND IMH FROM CLASSIC AORTIC DISSECTION A total of 529 imaging studies (CT, MR imaging, angiography, and TEES) were carefully re-reviewed on 214 patients originally classified as having an aortic dissection (87 type A and 127 type B). These studies were scrutinized with a radiologist specifically trained in vascular imaging. Among these patients, 36 were found on re-review to meet radiographic or morphologic criteria (described below) for the aortic dissections variants of PAU (19 patients) or IMH (17 patients). These 19 patients with PAU and 17 patients with IMH were then compared statistically to the remaining patients with classic aortic dissections (80 type A and 98 type B remaining after re-review).
Imaging Results The re-review of imaging studies initially classified as aortic dissection did indeed confirm the existence of radiologically distinct entities of PAU and IMH. Of the 212 patients
with an initial diagnosis of an aortic dissection, 36 patients met the criteria for clinical variants-19 with PAU and 17 with IMH. PAU and IMH were found to be more focal lesions than classic aortic dissection, which frequently propagates for much or the entire extent of the thoracoabdominal aorta. All patients classified as having PAU showed similar radiographic findings to the exemplary cases presented in Figures 7-10. All had contrast-filled outpouchings and an ulcerated crater seen on angiography, TEE, CT, or MR imaging. Associated intramural hematoma and pleural effusions were often identified as well. Severe aortic arteriosclerosis and calcification were seen uniformly. Patients diagnosed with IMH had imaging studies consistent with the representative examples of CT and TEE in IMH presented in Figures 11 and 12. There was absence of a flap traversing the aortic lumen, and imaging studies demonstrated the concentric orientation of the blood-containing space around the aortic lumen. These imaging studies were in distinct contrast with the findings in classic aortic dissection presented in Figures 4-6. The absence of an ulcer crater distinguishes these images of IMH from PAU with intramural hematoma.
646
COADY et a1
Figure 10. A, MR image (sagittal view) of a patient with a penetrating atherosclerotic ulcer. The penetrating atherosclerotic ulcer is diagnosed by visualization of a distinct ulcer crater in the absence of an intimal flap or false lumen. 6,MR image (axial view) of a patient with a penetrating atherosclerotic ulcer.
A number of different imaging modalities were employed to establish the diagnosis and to achieve follow-up in patients with PAU and IMH; more than one method was used for every patient (Tables 1 and 2). CT was the most commonly employed imaging modality. CT was used in 17 patients with PAU and 17 patients with IMH (94.4% of all patients). MR imaging studies were employed in 10 PAU and 4 IMH patients (38.8% of all patients). TEE was obtained in 15 PAU patients and 15 IMH patients (83.3% of all patients). Aortic
angiography was employed in 3 PAU patients and 6 patients with IMH (25.0% of all patients).
Demographic Considerations and Presenting Clinical Features
Tables 1 and 2 summarize the demographic and clinical features of the 19 PAU and the 17 IMH patients. Table 3 presents age and other
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
Figure 11. CT scan of a patient with an intramural hematoma. Note the presence of a concentric aortic wall hematoma and absence of an intimal dissection flap or penetrating ulceration.
Figure 12. TEE of a patient with an intramural hematoma. Note the absence of intimal disruption or penetrating atherosclerotic ulceration.
647
Moderate None Moderate None Mild None
F
75 69 69
56 69 64
14 15 16
17 18 19
None None Moderate
Mild None None None None Moderate None Moderate None
None None None Moderate None None Moderate Moderate None
Severe Moderate None None Moderate None Severe None None
None None
None Moderate
None Moderate None
Severe Moderate None
Moderate Moderate None Moderate Severe Moderate Moderate None None
Moderate Severe
Severe None
Renal Pulmonary Dysfunction Disease
-
Midscapular pain Mild left chest and midscapular pain Midscapular pain Left chest pain and hypotension Left chest pain Asymptomatic Midscapular pain Left chest pain Asymptomatic Asymptomatic Left chest pain Midscapular pain Anterior chest pain Midscapular pain Substemal pain Epigastric and chest pain Midscapular pain Midscapular pain Midscapular pain
Presenting Symptom
no no no
no no Yes
Yes Yes Yes no no Yes no no no
Yes no
Yes Yes
AAA
+ +
+ + + + ++ + ++ + + + + + +
+ + + + + + + + +
-
-
-
+ +
-
+ +
+ +
CT MRI TEE Angio
Management
Desc Desc Desc
Desc Desc Desc
Arch/Desc Asc Desc Desc Desc Desc Desc Desc Asc
Desc Desc
Surg Surg Surg
Surg Med Surg
Med Med Med Surg Med Med Med Surg Surg
Med Med
Desc Med Arch/Desc Med
Site
Rupture / Discharged Discharged Discharged
Rupture / Discharged Discharged Rupture / Discharged
Discharged Discharged Discharged Rupture / Discharged Discharged Discharged Discharged Discharged Rupture/Death
Rupture/Death Rupture/Death
Discharged Rupture/Death
Outcome
HTN = Hypertension; CT = computerized tomography; MRI = magnetic resonance imaging; TEE = transesophageal echocardiography; F = female; M = male; Desc = descending thoracic aorta; Asc = ascending aorta. Risk factors graded as per methods section. + Indicates study was performed for either diagnosis or follow-up. - Indicates study not performed. Med = medical management with antihypertensives and afterload reduction; Surg = patient underwent an operation. "Discharged implies that the patient was treated successfully and released from the hospital. "Rupture/Death" implies that the patient had an acute rupture and died in the hospital. "Rupture/Discharge" implies that the patient had an acute rupture prior to surgery, underwent resection of the penetrating ulcer, and was later released from the hospital.
F
F M
M F
Moderate Moderate Severe Moderate Severe Moderate Severe Severe Moderate
M M M F F F M F F
63 82 64 72 83 77 79 82 84
5 6 7 8 9 10 11 12 13
Mild None
Mild Severe
M M
76 85
3 4
Moderate None Severe Mild
Cardiac Disease
F M
Htn
77 79
Sex
1 2
Patient Age
Table 1. DEMOGRAPHIC AND CLINICAL CHARACTERISTICS OF PATIENTS WITH PENETRATING ATHEROSCLEROTIC ULCERS OF THE THORACIC AORTA
F F F
M MF
87 63
67 68 69 82 73 66
68 70
65 82
86
5
6
7 8 9 10 11 12
13 14
15 16
17
None
F
None
None Moderate
Severe None
None None Moderate Severe None None
Moderate Severe
None None Moderate None
Cardiac Disease
None
None None
Severe None
None None None Mild Mild None
Mild Severe
None None None None
None
None None
Mild Severe
Severe None Mild Moderate None Severe
None Mild
None Severe None None
Renal Pulmonary Dysfunction Disease
Midscapular pain Substemal pain Midscapular pain Chest pain and respiratory distress Intrascapular pain Anterior chest pain/ confusion Anterior chest pain Midscapular pain Midscapular pain Midscapular pain Anterior chest pain Substemal chest pain Lower back pain Substemal chest pain Midscapular pain Substernal chest pain Substemal chest pain
Presenting Symptom
n
n n
n n
o
o o
o o
yes n o n o yes no n o
yes n o
n o yes yes n o
AAA
+ +
+
+
+
+
CT
+
+ +
+ +
+
+ +
+
+
+
+ +
-
-
-
-
-
+
-
-
-
+
-
-
MRI
+
+
+ +
+ +
+
+
+ +
+
+
+
+
+ -
TEE
-
Asc
Surg
Desc Surg Asc Surg
-
-
Surg Med
Surg Med Med Med Med Surg
Med Med
Med Med Med Surg
Management
Desc Desc
Asc Desc Desc Desc Desc Asc
Desc Desc
Desc Desc Desc Asc
Site
-
-
+ + + -
-
-
+
-
-
+
Angio
Rupture / Death
Discharged Rupture/Discharged
Rupture / Discharged Discharged
Rupture / Discharged Discharged Discharged Discharged Discharged Death
Discharged Rupture/ Death
Discharged Discharged Discharged Rupture / Death
Outcome
HTN = Hypertension; CT = computerized tomography; MRl = magnetic resonance imaging; TEE = transesophageal echocardiography; F = female; M = male; Desc = descending thoracic aorta; Asc = ascending aorta. Risk factors graded as per methods section. + Indicates study was performed for either diagnosis or follow-up. - Indicates study not performed. Med = medical management with antihypertensives and afterload reduction; Surg = patient underwent an operation. "Discharged implies that the patient was treated successfully and released from the hospital. "Rupture/Death" implies that the patient had an acute rupture and died in the hospital. "Rupture/Discharge" implies that the patient had an acute rupture prior to surgery, underwent resection of the penetrating ulcer, and was later released from the hospital.
Mild Moderate
Moderate Moderate
Moderate Moderate None Moderate Mild Moderate
F M
M M
F F
M F F M
Moderate Moderate Severe Moderate
F
84 77 74 77
1 2 3 4
Moderate Severe
Htn
Sex
Age
Patient
Table 2. DEMOGRAPHIC AND CLINICAL CHARACTERISTICS OF PATIENTS WITH INTRAMURAL HEMATOMA OF THE THORACIC AORTA
650
COADY et a1
Table 3. COMPARISON OF TYPE A AND B AORTIC DISSECTIONS WITH INTRAMURAL HEMATOMAS AND PENETRATING ULCERS ~~
Type A Dissection
Patients Age Initial Aortic Size Prior AAA Rupture
80 (48 M, 32 F)
56.5 years ? 13.8 5.38 cm f 1.3 8 (10.0%) 6 (7.5%)
~
Type B Dissection
Penetrating Ulcer
Intramural Hematoma
98 (68 M, 30 F) 69.9 years* f 14.1 5.1 cm ? 1.3 30 (30.6%)$ 4 (4.1%)
19 (9 M, 10 F) 73.9 years* f 8.3 6.2 cmt ? 1.5 8 (42.1%)$ 8 (42.1%0)§
17 (7 M, 10 F) 74.0 years* f 7.8 5.5 cm -C 1.6 5 (29.4%) 6 (35.3%)§
Table reports means f standard deviation. Statistical analysis based on ANOVA using Bonferroni's method for multiple comparisons. = male; F = female. *The mean age for patients with type B dissection, penetrating ulcer, and intramural hematoma is significantly older than that of patients with type A dissection (P = 0.0001). tThe mean aortic diameter at initial presentation for patients with penetrating ulcer is significantly larger than for patients with type B dissection (P = 0.01). $The incidence of prior AAA for patients with penetrating ulcer is significantly greater than patients with type A dissection (P = 0.05). The incidence of prior AAA for patients with type B dissection is significantly greater than for patients with type A dissection M
(P =
0.01).
§The rupture rate for patients with penetrating ulcer is significantly higher than that of type A or B dissections ( P = 0.0001). The rupture rate for patients with intramural hematoma is significantly higher than either type A (P = 0.01) or type B dissection ( P = 0.001).
demographic and clinical data on the four categories of patients. Comorbidity
Accompanying medical problems were common in both PAU and IMH patients. For PAU, comorbidities included hypertension (18 patients, 94.7%), chronic obstructive pulmonary disease (12 patients, 63.1%), cardiac disease (8 patients, 42.1%), chronic renal insufficiency (5 patients, 26.3%), and diabetes mellitus (3 patients, 15.7%) (see Table 1).For IMH patients, comorbidities included hypertension (15 patients, 88.2%), chronic obstructive pulmonary disease (8 patients, 47.1%), cardiac disease (7 patients, 41.2%), and chronic renal insufficiency (5 patients, 29.4%) (Table 2). Pain Presentation
PAU and IMH patients presented almost uniformly with excruciating chest or back pain, similar to that of aortic dissection. As with classic dissection, anterior chest pain correlated with ascending aortic events and back pain with descending aortic events (see Tables 1 and 2). Ischemic Complications
No PAU or IMH patients manifested the branch vessel occlusion or run-off organ is-
chemia seen commonly with classic dissection.
Age Profiles
A distinct difference in the age profile was seen between classic dissection patients and those with PAU or IMH (see Table 3). Type B dissectors, PAU patients, and IMH patients were significantly older (means, 69.9, 73.9, and 74.0 years, respectively) than type A dissectors (mean 56.5 years) ( P = 0.0001). Although not statistically significant, patients with PAU and IMH tend to be older than those with type B dissection. Furthermore, PAU and IMH occurred within a narrow age range of elderly patients (63-85 and 63-87 years, respectively) (Fig. 13).
Aortic Sizes
All four aortic pathologies occurred in enlarged aortas (mean 5.3 cm). There were differences in aortic size at the time of occurrence of the various pathologic lesions. Patients with type A dissection, type B dissection, and IMH had similar initial aortic diameters (means 5.4, 5.1, and 5.5 cm, respectively). Patients with PAU had the largest aortas at presentation, measuring 6.2 cm (when compared to patients with type B dissection, P = 0.01) (see Table 3).
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
25
651
Type A Dissection
1
2ol
3 20-
8 &
. c) I
LI
0
z E 2
155 1
1
10-
5-
o', 20
I
30
I
I
I
40
B
I
I
50 Ages
I
60
I
I
70
I
I
I
80
Figure 13. A, Comparison of age distributions for type A dissections, penetrating atherosclerotic ulcer, and intramural hematoma. 6,Comparison of age distributions for type B dissections, penetrating atherosclerotic ulcer, and intramural hematoma.
Concomitant Abdominal Aortic Aneurysm
Dramatic differences were found in the incidence of concomitant abdominal aortic aneurysms (AAA). Patients with type B dissections, PAU, and IMH had much higher incidence of concomitant AAA (30.6%, P = 0.01,42.1%, P = 0.05, and 29.4%, respectively) when compared to type A dissection (10Y0). Patients with PAU had the highest incidence of accompanying AAA (42.1%) (see Table 3). Location
PAU was almost entirely a descending aortic phenomenon, with 17 of 19 (89.5%) occurring in the descending aorta. In 2 of these 17 patients, the hematoma extended subadventitially proximal to the ligamentum arteriosum into the aortic arch. The remaining 2 patients (10.5%) had ulcers located in the ascending aorta. IMH was also largely a descending aortic phenomenon, with 12 of 17
(70.6%) hematomas located in the descending aorta. Hospital Course
Figure 13 presents the incidence of rupture for all four lesions: type A dissection, type B dissection, PAU, and IMH. PAU were found to rupture much more commonly than either classic type of dissection (P = 0.0001). Similarly, IMH were found to rupture much more commonly than type A (P = 0.01) or type B dissection (P = 0.001) (Figure 14, Table 3). The incidence of acute rupture for PAU is 42.1% (8 of 19 patients) as opposed to 35.3% for IMH (6 of 17 patients), 7.5% (6 of 80 patients) for type A dissections, and 4.1% (4 of 98 patients) for type B dissections (P = 0.0001). All ruptures occurred with initial presentation or during initial hospitalization. With only one exception, all patients with PAU and IMH in the ascending aorta ruptured (see Tables 1 and 2). No ruptures occurred in late follow-up.
652
COADY et a1
(8/19)
(6/17)
(6/80)
Type A
Type B
Penetrating Ulcer
Intramural Hematoma
Figure 14. Rupture rates for type A dissections, type B dissections, penetrating atherosclerotic ulcer, and intramural hematoma.
Of the PAU patients, 8 of 19 (42.1%) ruptured prior to surgical treatment. Three individuals (Patients 2, 3, and 4 in Table 1) were thought initially to have a type B aortic dissection and died while being treated with standard medical therapy (P-blockade and afterload reduction) in the intensive care unit. Confirmatory autopsies were obtained in these three patients. The other 5 patients who ruptured and 3 additional patients without evidence of rupture (Patients 8, 12, 13, 14, 16, 17, 18, 19 in Table 1) were taken urgently or emergently to the operating room, with 7 (87.5%)surviving to discharge. The remaining 8 medically treated patients were discharged from the hospital following medical therapy. Of the IMH patients, 6 of 17 (Patients 4, 6, 7, 13, 16, and 17) (35.3%) ruptured prior to surgical treatment. An additional two IMH patients demonstrated expansion of their hematoma during the initial hospitalization (Pa-
tients 4 and 6, Table 2). Therefore, 8 of 17 IMH patients (47.1%) behaved malignantly during the initial presentation. Seven of these 8 patients underwent operations urgently or emergently, with 4 (50%) surviving to discharge; the eighth patient died of rupture prior to surgery. Pathology
In PAU, the gross pathology was confirmatory of surgical and radiographic findings. The aorta appeared densely atherosclerotic. The microscopic findings demonstrated intima1 degeneration and replacement with cholesterol clefts, which burrowed thorough the aortic media (Color Fig. 15). Depending upon the depth of penetration in specific patients, blood was found at varying levels within the media or beyond. The ulcer at times precipitated a localized intramedial dissection asso-
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
ciated with a variable amount of hematoma within the aortic wall. In IMH, gross findings in the operating room included a very superficial location of the hematoma in the aortic wall. In the operating room, in contradistinction to classic dissection, after the external adventitial layer was opened, the inner layer remained convex and appropriate in shape, not falling away from the aortic wall in concave collapse as seen in typical dissection. The IMH pathology uniformly demonstrated a significant degeneration of the aortic media, with hematoma located just cells away from the adventitia (Color Fig. 16). This finding is in contradistinction to our usual pathology findings in classic dissection, where the plane of hematoma is routinely found not more than two-thirds of the way through the media (see Figure 1).
653
radiologic follow-up, 3 individuals (Patients 8, 11, 14, Table 2) showed involution of their hematoma, 1 progressed to typical dissection with an intimal flap (Patient 3, Table 2) and 3 demonstrated no change in their hematomas over time (Patients 5, 9, and 10). All surgically treated patients surviving to discharge are well with no further aortic complications. Figure 17 displays the survival curves for all 19 PAU and all 17 IMH patients, compared to survival for the more common type A and type B dissectors. IMH patients had the lowest 1-year survival (65.8%) when compared to type A (88.9%) and type B (90.7%) dissectors ( P = 0.03). Although not achieving statistical significance, the 1-year survival for PAU patients (76.4%) falls short of the survival for both type A and type B dissectors. SUMMARY
Late Follow-up
Of the 4 surgical and 8 medically treated PAU patients surviving to discharge, all remain alive at mean follow-up of 31.5 months (range 3 to 84 months). On serial imaging, one individual (Patient 15, Table 1) has shown reduction in the size of his intramural hematoma. Of the 4 surgical and 9 medical IMH patients surviving to discharge, 2 died of unrelated causes (Patients 1 and 2, Table 2), and the others remain alive at mean follow-up of 26.4 months (range 1to 52 months). On serial
This article confirms the existence of two variants of acute aortic pathology, the penetrating atherosclerotic ulcer (PAU) and the intramural hematoma (IMH), which are radiologically distinct from classic aortic dissection. Table 4 reviews the characteristicsdistinguishing PAU from classic aortic dissection and IMH. We took as a matter of definition that classic aortic dissection involves a flap which traverses the aortic lumen. We defined PAU and IMH as nonflap lesions, with PAU demonstrating a crater extending from the
* Intramural Hematoma -
h
m
0.6 -
$
0.4
.-5
-
v)
0
3
6
9
12
Months
Figure 17. Kaplan-Meier 1-year survival for patients with penetrating atherosclerotic ulcer, intramural hematoma, type A, and type B aortic dissections. Asterisk indicates that P = 0.03.
654
COADY et a1
Table 4. CHARACTERISTICS DISTINGUISHING PENETRATING ULCER AND INTRAMURAL HEMATOMA FROM AORTIC DISSECTION Characteristic
Penetrating Ulcer
Intramural Hematoma
No intimal flap Intramural hematoma Localized ulceration penetrating internal elastic lamina Elderly hypertensive Hypertensive
No intimal flap No penetrating ulceration Hematoma within aortic wall
Intimal flap with contrast filling false lumen
Elderly Hypertensive
Same
Signs
Severe chest or midscapular pain Absent No branch vessel occlusion
Younger Hypertensive Occasionally bicuspid aortic valve Marfan’s syndrome Same
Absent No branch vessel occlusion
Location
Largely descending
Largely descending
Extent of lesion Degree of atherosclerosis Aortic size Concomitant AAA
Focal Always severe Largest (-6.2 cm) Common
Focal Variable Larger (-5.5 cm) Common
Diagnostic features
Typical patient
Symptoms
Aortic Dissection
Aortic insufficiency (type A) Pulse inequality Neurologic deficits compromise of blood flow to branch vessels Even distribution of ascending and descending Usually extensive Variable (often minimal) Smaller (-5.2 cm) Occasional (except type B, common)
AAA = abnormal aortic aneurysm.
aortic lumen into the space surrounding the aortic lumen. This categorization can be summarized with the expression, ”no flap, no dissection.” With these definitions made, re-review of the imaging studies for the present report identified 36 such lesions out of 214 cases originally read as aortic dissection. Therefore, these variant lesions accounted for over 1 out of 8 acute aortic pathologies. Besides confirming the existence of the conditions, PAU and IMH, as distinct radiographic lesions, this series strongly suggests that these two conditions constitute distinct clinical entities as well. Table 4 summarizes the clinical patterns of these two entities as apparent from the present study, and contrasts them with classic aortic dissections. In particular, the following observations, some of which are consonant findings in smaller series,+ can be made regarding the typical patient profiles of PAU and IMH from the present study: *References 1, 4, 11, 14, 15, 19, 23, 28
The patients with PAU and IMH are distinctly older than those with type A aortic dissection (74.0 and 73.9 versus 56.5 years, P = 0.0001). Although not statistically significant, PAU and IMH patients tend to be older than patients with type B aortic dissections as well. For PAU and IMH, unlike aortic dissection, the concentration in the elderly is manifested in a very small standard deviation of the mean age (see Fig. 13); these two entities, PAU and IMH, are essentially diseases of the seventh, eighth, and ninth decades of life. Patients with PAU and IMH are almost invariably hypertensive (about 94% of cases). The pain of PAU and IMH mimics that of classic aortic dissection, with anterior symptoms in the ascending aortic lesions and intrascapular or back pain with descending aortic lesions. Unlike classic dissection, PAU and IMH do not produce branch vessel compro-
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS
mise or occlusion and do not result in ischemic manifestations in the extremities or visceral organs. PAU and IMH are more focal lesions than classic aortic dissection, which frequently propagates for much or the entire extent of the thoracoabdominal aorta. PAU is uniformly associated with severe aortic arteriosclerosis and calcification, whereas classic dissection often occurs in aortas with minimal arteriosclerosis and calcification. PAU and IMH tend to occur in even larger aortas than classic aortic dissection (6.2 and 5.5 versus 5.2 cm, P = 0.01). PAU and IMH are strongly associated with AAA, which is seen concomitantly in 42.1% of PAU patients and 29.4% of IMH patients. PAU and IMH are largely diseases of the descending aorta (90% for PAU and 71% for IMH). Although our pathology data is limited, we do feel that an inherent difference in the histologic intramural level of the hematoma may underlie the pathophysiologic process that determines which patient develops a typical dissection and which develops an intramural hematoma. In particular, we feel that the level of blood collection is more superficial, closer to the adventitia, in IMH than in typical aortic dissection. This may explain why the inner layer does not prolapse into the aorta on imaging studies or when the aorta is opened in the operating room. This more superficial location would also explain the high rupture rates as compared to classic aortic dissection (Fig. 14, Table 3). We did find PAU and IMH to behave much more malignantly than typical descending aortic dissection. As seen in Figure 6, the rupture rate is much higher than for aortic dissection. Documented rupture occurred in 35.3% of IMH patients and in 42.1% of PAU patients. All but one PAU or IMH present in the ascending aorta ruptured. These ruptures all occurred during the initial hospitalization. These rupture rates are much higher than the 7.5% seen for ascending dissection and the 4.1% seen for descending dissection (P = 0.0001). (Of course, we do not ordinarily ob-
655
serve ascending dissections without surgery long enough to allow them to manifest their known poor prognosis). Therefore, PAU and IMH are extremely malignant in their early behavior. For this reason, it is essential to differentiate PAU and IMH from typical aortic dissection. Once patients with PAU and IMH survive the initial hospitalization, they do well, without late symptom recurrence or rupture in follow-up (mean 28 months in our study). What therapeutic correlates can we draw from this experience? First, detection is important. IMH and PAU are life-threatening conditions, which may not be detected on presentation unless the clinician is aware of the distinguishing radiologic features. These are enumerated in the previous paragraphs. Second, the following suggestions are made regarding management: We recommend early operative intervention for all PAU and IMH affecting the ascending aorta. Ascending lesions, although uncommon, in our experience uniformly ruptured or progressed and were often lethal. Therefore, these patients may be treated as those with classic ascending aortic dissections. This posture is supported by other series.'" 17, 28 Patients with PAU and IMH lesions in the descending aorta can be observed initially and treated with "antiimpulse" therapy, including P-blockade and afterload reduction. These lesions are more serious than classic descending aortic dissection, and a low threshold for surgical intervention must be maintained. If radiographic findings are ominous (severely bulging hematoma, extensive subadventitial spread, extra-adventitial blood, bloody pleural effusion, or deeply penetrating ulcer), surgery should be performed preemptively. In the absence of ominous initial radiographic findings, repeat imaging should be carried out within 3 to 5 days. In case of any progression, in volume or extent of the hematoma or overall size of the aorta, surgery should be performed preemptively to prevent rupture. Likewise, if symptoms are not controlled or recur on medical treatment, surgery
656
COADY et a1
should be performed. This posture is supported by most other series.14,15, 23, 28 If patients tolerate early medical management without clinical deterioration, our experience suggests that they may continue to be followed conservatively. In long-term follow-up, general size criteria for surgical intervention on the thoracic aorta can be a ~ p l i e d . ~ Kouchoukos and Dougenis recently reviewed strategies for evaluation and treatment of aortic pathologies.12Our findings and clinical recommendations are consonant with those reported in smaller series of variant aortic pathologies examined in that review. As IMH and PAU primarily affect older individuals with significant comorbidity, however, our above recommendations must be tempered by consideration of the patient’s overall condition and general outlook. Some of these patients may not be appropriate candidates for the aggressive surgical strategy outlined. We are just beginning to understand the identity and behavior of these lesions, and our management suggestions are provisional recommendations regarding the treatment of penetrating ulcer and intramural hematoma. Continued experience will be required to assess more completely the natural history of these variants of aortic dissection and to achieve more definitive treatment recommendations. Furthermore, although we have described these variant lesions as entities distinct from aortic dissection, it is likely that intermediate lesions exist, and that ulcer, hematoma, and dissection lie on a continuum of disease processes. References 1. Alfonsa F, Goicolea J, Aragoncillo P, et al: Diagnosis of aortic intramural hematoma by intravascular ultrasound imaging. Am J Cardiology 76:735-738, 1995 2. Coady MA, Rizzo JA, Hammond GL, et al: Penetrating ulcer of the thoracic aorta: What is it? How do we recognize it? How do we manage it? J Vasc Surg 271006-1016, 1998 3. Coady MA, Rizzo JA, Hammond GL, et al: What is the appropriate size criterion for resection of thoracic aortic aneurysms? J Thorac Cardiovasc Surg 113:47& 491, 1997
4. Cooke JP, Kazmier FJ, Orszulak TA: The penetrating aortic ulcer: Pathologic manifestations, diagnosis, and management. Mayo Clin Proc 63:718-725, 1988 5. Crawford ES, Svensson LG, Coselli JS, et al: Aortic dissection and dissecting aortic aneurysms. Ann Surg 208(3):254-273, 1988 6. Gore I: Pathogenesis of dissecting aneurysms of the aorta. Arch Pathol Lab Med 53:142-153, 1952 7. Gore I, Seiwert VJ: Dissecting aneurysm of the aorta. Arch Pathol 53:121, 1952 8. Hirst AE, Gore I: Is cystic medionecrosis the cause of dissecting aortic aneurysm? (editorial). Circulation 53(6):915-916, 1976 9. Hirst AE, Gore I: The etiology and pathology of aortic dissection. In Doroghazi RM, Slater EE (eds): Aortic Dissection. New York, McGraw-Hill, 1983 10. Hirst AE Jr, Johns VJ Jr, Klime SW Sr: Dissecting aneurysm of the aorta: A review of 505 cases. Medicine (Baltimore) 37217-279, 1958 11. Kazerooni EA, Bree RL, Williams DM: Penetrating atherosclerotic ulcers of the descending thoracic aorta: Evaluation with CT and distinction from aortic dissection. Radiology 183:759-765, 1992 12. Kouchoukos NT, Dougenis D. Surgery of the thoracic aorta. NEJM 26:1876-1888, 1997 13. Larson EW, Edwards WD: Risk factors for aortic dissection: a necropsy study of 161 cases. Am J Cardiol 53(6):849-855, 1984 14. Mozsowitz HD, Lampert C, Jacobs LE, et al: Penetrating atherosclerotic aortic ulcers. Am Heart J 128:1210-1217, 1994 15. Muluk SC, Kaufman JA, Torchiana DF, et al: Diagnosis and treatment of thoracic aortic intramural hematoma. J Vasc Surg 24:1022-1029, 1996 16. Murray CA, Edwards JE: Spontaneous laceration of the ascending aorta. Circulation 472348-858, 1973 17. Murray JG, Manisali M, Flamm SD, et al: Intramural hematoma of the thoracic aorta: MR image findings and their prognostic implications. Radiology 204:349355, 1997 18. Nakashima Y, Kurozami T, Sueishi K, et al: Dissecting aneurysm: A clinicopathologic and histopathologic study of 111 autopsied cases. Hum Pathol 21L:291296, 1990 19. Neinaber CA, von Kodolitsh Y, Peterson B, et al: Intramural hemorrhage of the aorta: Diagnostic and therapeutic implications. Circulation 92:1465-1472, 1995 20. Patterson DL, Brennan S, Cartwright T, et al: Traumatic rupture of an aortic ulcerative atherosclerotic plaque producing aortic dissection: A complication of interscapular back blows used to dislodge objects from the esophagus. Clinics in Cardiology 16:741744, 1993 21. Roberts WC: Aortic dissection: Anatomy, consequences, and causes. Am Heart J 101:195-214, 1981 22. Roberts WC, Honig HS: The spectrum of cardiovascular disease in the Marfan syndrome: A clinicomorphologic study of 18 necropsy patients and comparison to 151 previously reported necropsy patients. Am Heart J 104:115, 1982 23. Schappert T, Sadony V, Schoen F, et al: Diagnosis and therapeutic consequences of intramural aortic hematoma. J Card Surg 9:50&515, 1994 24. Schnitker MA, Bayer CA: Dissecting aneurysm of aorta in young individuals, particularly in associated with pregnancy. Ann Intern Med 20:486, 1944 25. Shachter N, Perloff JK, Mulder DG: Aortic dissection in Noonan’s syndrome (46 XY Turner). Am J Cardiol 54(3):464-465, 1984
PATHOLOGIC VARIANTS OF THORACIC AORTIC DISSECTIONS 26. Shennan T Dissecting aneurysms. Medical Research Council, Special Report Series, No. 193, 1934 27. Slater DN, Grundman MJ, Mitchell L: Turner’s syndrome associated with bicuspid aortic stenosis and dissecting aortic aneurysms. Postgrad Med J 58(681):436-438, 1982 28. Stanson AW, Kazmier FJ, Hollier LH, et al: Penetrating atherosclerotic ulcers of the thoracic aorta: Natural history and clinicopathological correlations. Ann Vasc Surg 1:15-23, 1986 29. Stolle CA, Pyeritz RE, Myers JC, et al: Synthesis of an altered type I11 procollagen in a patient with type IV Ehlers-Danlos syndrome: A structural change in the alpha 1 (111) chain which makes the protein more susceptible to proteinases. J Biol Chem 60:1937, 1985
657
30. Svensson LG, Crawford ES: Occlusive disease of the aorta. In Svensson LG, Crawford ES (eds): Cardiovascular and Vascular Disease of the Aorta. Philadelphia, WB Saunders Company, 1997 31. Wheat MW. Pathogenesis of aortic dissection. In Doroghazi RM, Slater EE (eds): Aortic Dissection. New York, McGraw-Hill, 1983, p 55 32. Wolinsky H: Response of the rat aortic media to hypertension: Morphological and chemical studies. Circ Res 26(4):507-522, 1970 33. Yamada T, Takamiya M, Naito H, et al: Diagnosis of aortic dissection without intimal rupture by x-ray computed tomography. Nippon Acta Radio1 45:699710, 1985
Address reprint requests to John A. Elefteriades, MD Section of Cardiothoracic Surgery Yale University School of Medicine 333 Cedar Street New Haven, CT 06510