Natural history and serial morphology of aortic intramural hematoma: A novel variant of aortic dissection Isidre Vilacosta, MD, a Jose Alberto San RomAn, MD, b Joaquin Ferreirbs, MD, c Pa loma Aragonci l lo , MD, d Ramiro M6ndez, MD, c Juan Anton io Castillo, MD, a Maria Jesfls Rollfin, MD, a Elena Batlle, MD, a Vicente Peral, MD, a and Luis Sf inchez-Harguindey, MD, a Madrid and Valladoli~ Spain

Background Acute aortic dissection is a cardiovascular emergency that requires prompt diagnosis and treatment. Transesophageal echocardiography is the current standard diagnostic imaging modality in many medical centers. Aortic intramural hematoma is a variant of aortic dissection whose natural history and prognosis have not been well studied. We performed transesophageal echocardiography in patients with aortic intramural hematoma to determine the echocardio- graphic characteristics and echocardiographic evolution of this lesion, impact on patient management, and patient outcome.

Methods and Results Twenty-one consecutive patients with aortic intramural hematoma confirmed anatomically (four patients) or with an additional diagnostic imaging technique (17 patients) underwent a transesophageal echocardiographic examination. Fifteen patients with longstanding hypertension had chest or back pain, and the intramural hematoma was visualized in the ascending aorta (n = 4), along the whole aorta (n = 4), in the descending aorta (n = 6), or in the aortic arch (n = 1 ). The thickening of the aortic wall was crescentic. Patients with ascending aortic intramural hematoma had the following results: two patients died suddenly, three patients underwent surgery because of increased aortic wall thickening (one patient) or secondary intimal tear (two patients), and the remaining three patients had regression of the hematoma. Patients with hematoma confined to the descending aorta and the patient with aortic arch involvement (n = 7) had a different result: one patient died from aortic rupture and the remaining six patients did well. Six patients had a traumatic aortic injury, and the intramural hematoma was located along the descending thoracic aorta. The thickening of the aortic wall was circular in five patients and crescentic in one. Three of these patients had normalized thickness of the aortic wall on follow-up transesophageal echocardiographic studies. The other three patients died from multiorgan system failure. Aortography showed a reduction of the diameter of the aortic lumen in four patients; diameter in the remaining 17 patients was normal.

Conclusions Aortic intramural hematoma can be detected and monitored by transesophageal echocardiography but not by aortography. Two types of aortic intramural hematoma can be distinguished: (1) traumatic of good prognosis and (2) nontraumatic, which can be an early stage of the classic aortic dissection, with bad prognosis in cases involving the ascending aorta. (Am Heart J 1997;134:495-507.)

Acute dissection of the aorta is one of the most

dramatic cardiovascular emergencies. To limit the

possibility of death, a detailed morphologic and

functional diagnosis must be quickly obtained.

From the aDepartments of Cardiology, CRadiology, and dHurnan Pathology, Hospital Universitarlo de San Cados, Madrid, and the bDepartment of Cardiology, Hospital Universitario de Valladolid. Received for publication Aug. 5, •996; accepted April 24, 199Z Reprint requests: Dn Isidre Vilacosta, Serrano 46, 28001 Madrid, Spain. Copyright ©) 1997 by Mosby- Year Book, Inc. 0002-8703/97/$5.00 + 0 4/1/83046

Most aortic dissections are characterized by a

transverse and longitudinal separat ion of the aortic

media and an int imomedial tear. 1,2 However, aortic

intramural hema toma (AIH) represents a variant of

d issec t ion character ized by the absence of an

intimal tear. 3 In this condi t ion, the false channe l is

p robab ly c r e a t e d by a hemor rhage of the aortic

vasa vasorum into the aortic wall. 4 This

noncommunica t ing form of aortic d issec t ion has

been increasingly repor ted with m o d e r n d iagnost ic

imaging techniques . 3,5-7 At present , the use of

4 9 6 Vilacosta et al.

Pt. Age (yr)/ Location of no sex AIH

(b)/major Ao (a) mm (b) mm

(c)/Ao ratio (c) mm

Type of Ao Ao ratio wall thickening

1 65/M Asc Ao + Asc: 32 13 45 Desc Ao Desc: 25 15 40

2 78/M Asc Ao 7 35 42

3 78/M Asc Ao 25 32 57

4 67/M Asc Ao + Asc: 22 34 56 Desc Ao Desc: 12 28 40

5 66/M Desc Ao 20 28 45 6 75/M Desc Ae 22 20 42

7 70/F Desc Ao 15 26 41 8 70/M Desc Ao 10 32 42

9 59/F Asc Ao + Asc: 6 32 38 Desc Ao Desc:6 28 34

10 65/M Desc Ao 15 17 32 11 79/M Asc Ao 12 28 40 12 72/M Asc Ao + Asc: 8 30 38

Desc Ao Desc: 15 16 31 13 82/F Ao Arch + Arch: 10 25 35

Desc Ao Desc: 10 22 32 14 4 2 / M Asc Ae 18 28 46 15 68/M Desc Ao 20 30 50

0.29 Asc: circular 0.38 Desc: crescentic

0.83 Crescentic

0.56 Crescentic

0.61 Asc: crescentic 0.70 Desc: crescentic

0.62 Circular, crescentic 0.48 Circular, crescentic

0.63 Crescentic 0.76 Crescentic

0.84 Crescentic 0.82 0.53 Crescentic 0.70 Crescentic 0.79 Crescentic 0.52 0.71 Crescentic 0.69 0.61 Crescentic 0.60 Circular

AO, Aorta; asc, ascending; desc, descending.

Pt. Age (yr)/ Location of no sex AIH (a)

Ao wall thickness (g)/minor Ao lumen (b)/major Ao (c)/Ao ratio

mm (b) mm (c) mm Type of Ao

Ao ratio wall thickening

1 30 /M Desc. aorta 2 32 /M Desc. aorta 3 29 /M Desc. aorta

4 76 /M Desc. aorta

5 6 9 / M Desc. aorta 6 6 3 / M Desc. aorta

11 20 31 0.65 Circular 19 20 39 0.51 Circular 6 30 36 0.83 circular

8 32 40 0.80 Circular

5 29 34 20 20 40

0.85 Circular 0.50 Crescentic

AO, Aorta; asc, ascending; desc, decending; IABP, interaaortic balloon pump.

t r ansesophagea l echoca rd iography (TEE) can be

cons idered the diagnost ic p rocedure of choice in

assessing pat ients with a suspec ted aortic

dissection. 8 AIH has scarcely been evalua ted by

TEE. 3 In this s tudy we descr ibe the echocard iographic criteria, clinical and echocard iographic evolut ion, and prognos is of

pat ients with AIH. Two types of AIH have been

dist inguished: traumatic, and nontraumatic .

M e t h o d s Patient population

Twenty-one patients seen between March 1989 and November 1995 formed our study group. This consecutive

American Heart Journal September 1997

Vilacosta et al. 497

Echo-free Flow spaces within AIH Confirmation

Clinical presentation

Clinical and Outcome/ echocardiographic follow-up

evolution time

Yes No MRI

Yes No CT, autopsy

Yes No CT, surgery

Yes No CT

Yes No CT + MRI Yes Yes CT + MRI

Yes No CT + MRI No No CT, surgery

Yes No CT + MRI

Yes Yes CT + MRI No No CT Yes Yes CT

No No MRI

No No MRI No No CT

Chest pain, paraplegia

Chest pain

Chest pain, syncope Chest pain

Chest pain Back pain,

Chest pain Chest pain

Chest pain, syncope Back pain Back pain Back pain

Chest pain

Chest pain Chest pain

Early partial regression, Alive/3 yr late intimal disruption (type II Ao dissection); no surgery

Adventitial rupture; Death/t6 hr sudden death

Progression of AIH; Alive/2 yr shock (surgery)

Late intimal disruption Death/20 d (Type I Ao dissection); syncope (surgery) No change Alive/8 mo Partial regression Alive/8 mo syncope No change Alive/2 yr Adventitial rupture Death/48 hr

(surgery) Complete regression Alive/8 mo

Partial regression Alive/10 mo Partial regression Alive/6 mo No change Death/24 hr Sudden death Complete regression Alive/4 mo

Complete regression Alive/11 mo Complete regression Alive/2 yr

Echo-free Flow spaces within AIH Confirmation

Clinical presentation

Clinical and Outcome/ echocardiographic follow-up

evolution time

No No CT Yes No CT No No CT, autopsy

No No CT

No No CT Yes Yes CT

Trauma Trauma Trauma failure (No change) Trauma (No change) Trauma IABP

Complete regression Alive/2 yr, 8 mo Complete regression Alive/2 yr, 6 mo Multiorgan system Death/36 hr

Renal failure and sepsis Death/24 d

Complete regression Alive/1 yr, 2 mo late intimal disruption Death/12 d (type III Ao dissection)

group of patients is a composite of two subgroups, a

subgroup of 15 (17%) was selected from a group of 88

patients diagnosed by TEE with aortic dissection from

nontraumatic sources. These 15 patients had signs and

symptoms consistent with acute aortic dissection; the classic

mobile dissection flap was not visualized and a thickened

aortic wall without an intimal tear in the TEE study was

detected. Consequently, they were categorized as having a

noncommunicating aortic dissection (group 1). In the same

period 22 patients admitted to our institutions with severe

blunt chest trauma with a widened mediastinal silhouette

were studied with TEE. Five patients (22%) had increased

aortic wall thickness. A dissection flap and an entrance tear

were not detected; and they were diagnosed as having an

American Heart journal Volume 134, Number 3

498 Vilacosta et al.

Left, Cross-sectional TEE view of AIH (asterisk) confined to descending thoracic aorta (,40). Right, Schematic depicting echocardiographic aortic measurements, a, Aortic wall thickness; b, minor aortic lumen diameter; c, major aortic diameter.

AIH. Four patients had a classic aortic dissection, and the

remaining 13 patients with chest trauma had a normal aorta

by TEE. During the study period, one patient had a

barotraumatic aortic injury from an intraaortic balloon pump

that was studied by TEE and diagnosed as AIH. These six

patients with AIH formed group 2.

Overall, 21 patients were considered to have an AIH: 15

nontraumatic (group 1) and six traumatic (group 2).

Confirmation of this echocardiographic diagnosis was

performed by surgery or necropsy (four patients) or with a n

additional imaging technique---computed tomography (CT)

or magnetic resonance imaging (MRI)--and on the basis of

clear follow-up TEE changes (17 patients). Conventional

aortography was done in all patients. Meticulous clinical and

echocardiographic follow-up was performed in all patients.

Group 1 was composed of 12 men and three women (aged

48 to 82 years). Group 2 was composed of six men (aged 29

to 76 years). Tables I and II describe clinical presentation

and patient demographics.

Echocardiographic evaluation All TEE examinations were performed with commercially

available 5 MHz monoplane, biplane, and multiplane

transducers (Toshiba SSH 160A, Toshiba SSH 140 A, and

ATL, Ultramark 9). Before the transesophageal examination,

patients were sedated with an intravenous short-acting

benzodiazepine w-ith the addition of meperidine (25 to 50

mg). Local anesthesia of the pharynx was achieved with

benzocaine 20% spray to suppress the gag reflex. TEE studies were safely completed within 15 + 6 minutes. The

TEE study included standard views of the heart followed by

a complete two-dimensional and color-flow mapping examination of the ascending, arch, and descending

MRI. Sagittal section shows AIH (arrows) extending throughout extent of descending thoracic aorta. There was no flow in thickened aortic wall.

thoracic aorta. The ascending part of the aortic arch could

not routinely be visualized because of interposition of the

left bronchus.

Classic aortic dissection was diagnosed when a consistent

mobile linear echo indicative of a dissection flap and an

intimomedial tear were seen within the aortic lumen. 8 Two

specific echocardiographic criteria were used to differentiate

AIH from classic aortic dissection in these patients with

clinically suspected acute aortic dissection: (1) circular or

crescentic thickening (>5 ram) of the aortic wall (this

thickened aorta had a thrombuslike echo pattern), and (2)

absence of an intimomedial tear. 3 Measurements of the aortic

diameter and wall thickness were performed at the level of

the maxi/nal thickness of the AIH from the intima to the

adventitia (Fig. 1). The longitudinal extension of the AIH was

evaluated by scanning the whole aorta from the aortic valve

to below the diaphragm. Several other echocardiographic

findings were also assessed: central displacement of intimal

calcification, a layered appearance of the thickened aortic

wall, and presence of echo-free spaces and color Doppler

flow signals within the AIH. All but three patients who died

within 48 hours from admission had between three and

seven TEE studies. During the echocardiographic follow-up studies six types of evolution of the AIH were expected: (1)

no change in wall thickness, (2) increase of wall thickness

(progression of AIH), (3) decrease of wall thickness (partial regression of AIH), (4) normalization of the thickened aortic

American Heart ,Journal September 1997

Vilacosta et al. 4 9 9

Autopsy specimen of AIH (asterisk) in ascending aorta. Aortic intima is intact. A, Transverse section; B, longitudinal section.

wall (total regression of AIH), (5) development of a typical

dissection (visualization of an entrance tear), and (6) rupture

of the adventitia of AIH. If the development of a classic

dissection occurred in the follow-up studies, the location and

extent of the dissection flap were used to determine the type

of aortic dissection according to DeBakey classification: type

I if the flap was seen in the ascending and descending

thoracic aorta, type II if the flap was confined to the

ascending aorta, and type III if the flap was seen only in the descending thoracic aorta.

To establish TEE criteria that would distinguish AIH from type III chronic aortic dissection with thrombosed false

lumen or from a descending aortic aneurysm with mural

thrombosis, six consecutive patients with a chronic aortic

American Heart JournaJ Volume 134, Number 3

500 Vilacosta et a].

Transesophageal longitudinal (left) and transverse (right) views of AIH (asterisk) located in ascending aorta (,40). AI, Left atrium.

dissection and a clotted false lumen and six patients with a

surgically proven aortic aneurysm with mural thrombus were

also enrolled into the study.

CT CT scans were obtained in 11 of the 15 patients in group 1

and in all patients in group 2. Intravenous boluses of 80 to

150 ml of nonionic contrast medium were used for radio-

graphic CT. Transverse CT slices of 10 mm were obtained

from the arch to the aortic bifurcation. AIH was defined by

cT as a nonenhancing circular or crescentic thickening of the

aortic wail without evidence of the typical intimal flap. 5

Central displacement of intimal calcification was specifically

sought. 5 The presence of mediastinal hemorrhage and

pleural effusion was also evaluated. CT studies were

reviewed independent of the TEE findings for the presence

and longitudinal extension of AIH.

MRI MRI scanning has the capability to estimate the age of

a thrombus by assessing methemoglobin versus oxyhemo-

globin spin-echo signals. By using the fat-suppression

technique, AIH was defined as a circular or crescentic high-

intensity thickening of the aortic wall without evidence of

the classic intimal flap (Fig. 2) and absence of blood flow

within the thickened aortic wall during cine MRI. 7

Transverse, coronal, and oblique sagittal planes were

obtained in all patients. Cine MRI was performed in all patients to assess flowing blood in the thickened aortic wall

segments. MRI studies were done in 8 of the 15 patient in

Transesophageal longitudinal views of AIH in ascending aorta (top arrows). Bottom, echo-free areas (arrows) located immediately beneath dissection flap can be seen. AI, Left atrium; AO, aorta; AP, pulmonary artery.

group 1. No patient from group 2 had an MRI study. All

MRI studies were reviewed independently.

Aortography Aortograms were performed in the standard manner in all

patients in both groups, and all were reviewed by

experienced angiographers who were unaware of the

echocardiographic findings. An angiographic diagnosis of

aortic dissection was established with the identification of an

intimal flap or a double lumen. Reduction of the diameter of

the aortic lumen and rectification of the aortic contour were

specifically sought.

Results Nontraumatic AIH

Table I p resen t s clinical character is t ics of the s tudy

pat ients . All pa t ien ts h a d a long h is tory of hype r t ens ion .

American Heart Journal September 1997

Vilacosla et al. SO1

A, Admission CT scan of patient 4, who was seen with acute chest pain. Mild crescentic thickening of descending thoracic aortic wall was seen. Severe paricardial effusion is also shown. B, Seventeen days later new CT scan displayed classic type I aortic dissection.

Eleven patients had chest pain and four had back pain

on admission and were clinically indistinguishable from

patients with classic aortic dissection. In addition to the

chest or back pain, three patients had syncope and one

had paraplegia at the time of admission. Renal function

was acutely compromised in two cases in whom AIH

involved the descending aorta (one from compromise

of the left renal artery). Chest radiograph on admission

revealed a widened mediastinum caused by a

prominent aortic silhouette in seven patients and

A, Cross-sectional TEE view of patient 1. Huge AIH (arrows) can be observed. Regression of aortic hematoma is demonstrated in B. Secondary intimal tear and formation of false channel (asterisk) is seen in C. AI, Atrium; AO, aorta.

pleural effusion in six. Aortic regurgitation was found in

three patients, and all had an AIH in the ascending

aorta. Diagnosis of AIH was confirmed by anatomy

(three patients) (Fig. 3), CT (four patients), MRI (three

American Heart journal Volume 134, Number 3

502 Vilacosta et al.

MRI scan of patient 14. Large intramural hematoma involving ascending aorta is present (asterisk) on admission (A). Three months later hematoma regressed (B).

patients), or CT and MRI (five patients). Aortography showed a reduction of the aortic lumen in two patients and rectification of the aortic contour in one patient; in the remaining 12 patients the aortogram was normal.

Echocardiography Table I summarizes echocardiographic findings. The

AIH was visualized in the ascending aorta in four

patients (Fig. 4), along the whole aorta in four

patients, in the descending aorta in six patients, and in /

the aortic arch in one patient. The maximal aortic wall

thickness at the site of t/he AIH ranged from 6 to 32 mm (mean 20 mm), the mean minor aortic lumen

diameter at the site of the AIH was 34 mm (range 13

to 35 mm), the maximal diameter of the aorta at the

level of AIH varied between 31.and 57 mm (mean 55

mm), and the mean aortic ratio (minor aortic lumen diameter divided by major aortic diameter) at the level

of AIH was 0.84 (range 0.29 to 0.84). In 11 patients the hematoma led to a crescentic aortic wall

thickening; in three patients the thickening was

crescentic in some aortic segments and circular in

others, and one patient had an asymmetrical circular

thickening of the aortic wall. In 10 of these 15 patients echo-free spaces within the AIH were present, and in

only three patients a very slow flow with color

Doppler flow mapping could be visualized within the

echolucent areas. In the remaining patients no flow could be detected within the AIH. In three patients

these echolucent areas were located immediately

beneath the intimomedial dissection flap, permitting

the visualization of the dissection flap (Fig. 5). Three

patients with involvement of the ascending aorta had pericardial effusion.

Clinical and echocardiographic evolution Medical therapy and close clinical and

echocardiographic follow-up was judged to be the

adequate treatment of choice for ail patients at the

time of admission. In eight patients the AIH affected

the ascending aorta (Table I). Two patients died suddenly in the first 24 hours after initial diagnosis (in

one patient AIH and adventitial rupture were

documented at necropsy), one patient underwent surgery because of an increase of the aortic wall thickening (progression of AIt-I) and hypotension, and

in one patient the initial AIH evolved into a type I classic dissection and required surgery (Fig. 6).

Another patient came to the emergency room with a

huge AIH (Fig. 7, A) and moderate aortic regurgitation; 36 hours later the AIH regressed

dramatically (Fig 7, B) and there was no aortic regurgitation. A routine TEE study several months later revealed an asymptomatic secondary intimal tear (Fig 7, C). In the remaining three patients with AIH

affecting the ascending aorta, regression of the intramural hematoma, partial (n = 1) and complete (n = 2), was documented (Fig. 8).

American Heart journal September 1997

Vilacosta et al. 503

AIH (n = 6)

Classic chronic Ao dissection with thrombosed

false lumen (n = 6)

M u r a l thrombosis in aortic

aneurysm (n = 6)

Presence of spontaneous contrast 1 {17%} Maximum thickness of aortic echodensity 17 mm Presence of dissection flap 0 Type of aortic echodensity (circular/crescentic) I /3 * Echo-free spaces 4 (67%) Presence of flow within the aortic echodensity 2 (33%) Minor aortic lumen diameter 28 mm Major aortic diameter 42 mm Aortic ratio: (aortic lumen/aortic diameter) 0.60 Presence of distinct echogenic wall layers 1 (16%) Central displacement of Ca++ 4 (66%)

2 (33%) 5 (83%) 27 mm 13 mm

5 (83%) 0 0/6 2/4

4 (67%) 1 (16%) 2 (33%) 0 25 mm 48 mm 40 mm 65 mm 0.66 0.76

2 (33%) 1 (16%) 4 (66%) 0

Ao, Aortic. *Two patients hod both types of thickening (circular and crescentic) along the descending aorta.

Cross-sectional TEE views of patient with AIH confined to descending thoracic aorta (left). One month later (right) regression of AIH is shown.

Cross-sectional TEE view of patient with traumatic AIH. Existence of different echogenic wall layers (arrows) can be observed. AO, aorta.

Six patients in w h o m AIH exclusively involved the descending aorta and in one patient in which the aortic arch was the main aortic segment affected had a different result (Table I). One patient died from aortic rupture, whereas the remaining six patients did well during their follow-up. Two patients' aortic wall thickness did not change, whereas in four patients

the thickened aortic wall normalized totally (two patients) or partially (two patients) (Fig. 9). The patient with aortic arch involvement (patient !3) had normalized aortic wall thickness during follow-up.

Six patients with a classic type III chronic aortic dissection with a clotted false lumen and six patients with a descending thoracic aortic aneurysm with mural thrombosis were enrolled into this study to be

American Heart Journal Volume ]34, Number 3

504 Vilacosta et a].

Transesophageal longitudinal (left) and transverse (right) views of AIH located in descending thoracic aorta. Central displacement of intimal calcification can be observed (arrows).

Top, Cross-sectional TEE view of patient with traumatic AIH (asterisk). Bottom, Normalization of aortic wall thickness was demonstrated several days later. AO, Aorta; DPL and PE, pleural effusion.

Traumatic A IH

Six patients had a traumatic aortic injury, five had had

a traffic accident and a widened mediastinal silhouette

on chest radiograph, and one patient had a barotrau-

matic aortic injury from an intraaortic bal loon pump

(Table I!). Diagnosis of AIH was confirmed by anatomy

in one patient and by CT in all patients. Aortography

was normal in four patients; in patient 2 a reduction of

the diameter of the aortic lumen could be recognized.

In patient 6 aortography was performed on follow-up.

echocardiographical ly compared with the six patients

with AIH confined to the descending aorta because

these three aortic pathologic conditions share a similar

echocardiographic image. Table III summarizes the

echocardiographic characteristics of these patients. The

classic dissection flap was seen in five patients (83%) with chronic aortic dissection; in this group four

patients had echo-free spaces and central displacement

of calcium. Five patients with a descending aortic

aneurysm had spontaneous contrast within the aortic

lumen; in this subgroup of patients the presence of a

dissection flap, visualization of flow within the aortic echodensity, and central displacement of calcium could not be detected.

Echocardiography

Table II summarizes echocardiographic findings. The

AIH was localized along the descending aorta with a

variable longitudinal extension. The maximal aortic wall

thickness at the site of AIH ranged from 5 to 22 mm

(mean 1i mm), the mean minor aortic lumen diameter

at the site of the MH was 25 mm (range 20 to 32 mm),

the maximal diameter of the aorta at the level of AIH

varied between 31 and 40 mm (mean 36 mm), and the

mean aortic ratio at the level of AIH was 0.69 (range

0.50 to 0.85). In five patients the AIH led to a circular

aortic wall thickening; the patient with an intraaortic

balloon pump had a crescentic type of wall thickening.

Echo-free spaces within the AIH could be visualized in

two patients. The dissection flap was seen in one

patient. Blood flow within the aortic wall thickening

American Heart Journal September 1997

Vilacosia et al. $05

could be only detected in the patient with a

barotraumatic aortic injury. In two patients shearing of

different echogenic wall layers during the cardiac cycle

could be observed (Fig. 10).

Clinical and echocardiographic evolution As for the nontraumatic group, a conservative

medical approach was chosen. Three patients had

normalized aortic wall thickness (complete regression,

Fig. 11) during follow-up and did well. Two patients

died early in their follow-up, one from a multiorgan

system failure and the other fi'om septicemia and

renal failure. In these two patients changes in aortic

wall thickness could not be demonstrated. The pat ient

with a barotraumatic aortic injury that resulted in a

classic type III aortic dissection confirmed by an

aortography eventually died from septicemia and

multiorganic failure. 9

Discussion Most aortic dissections are characterized by a key

finding: an intimomedial tear. 1 This inner wall laceration

is considered to be the beginning of the dissection and

has been called the entrance tear.: Thereafter blood

under pressure dissects the media longitudinally, and as

a consequence a double-channel aorta is formed. The

partition between the true and false channels is the so-

called intimomedial flap. 2 Necropsy data have showed

that in <5% of patients with dissection it is not possible

to locate an entry tear. :° In the clinical setting, some

authors have also noted the absence of an entrance tear

in a small[ proportion of patients with dissection studied

with noninvasive imaging techniques (TEE, CT, and MRI).3'5-7'11-13 This variant of dissection (noncom-

municating aortic dissection) has been denominated

MH. 3 In these cases, the false lumen is created by a

hemorrhage into the aortic media, most likely after

rhexis of the vasa vasorum that penetrate the outer half

of the media from the adventitia and arborize at this

level. 4 The natural history, treatment, and prognosis of

MH have not been well delineated. TEE has excellent

sensitivity in diagnosing aortic dissection, can be

quickly performed at the patient's bedside, and is

readily available in most hospitals. Therefore many

authors maintain that TEE should be considered the first

diagnostic technique of choice in cases of suspected

aortic dissection. 8 In this study, we have shown the

echocardiographic evolution, natural history, and

prognosis of two types of MH (traumatic and

Schematic drawing of evolution of patient 1. Hematoma formation, regression, and late intimal disruption. AI, Left atrium; AO, aorta; VI, left ventricle.

nontraumatic), with TEE as the first-line imaging modality for diagnosing aortic dissection.

Diagnosis of AIH Nontraumatic AIH affected patients with longstanding

hypertension who were clinically undistinguishable from patients with classic aortic dissection. In our study comprising a large number of patients, the frequency of nontraumatic AIH was 17%. In other clinical series, noninvasive imaging techniques identified AIH in 12.8% (MRI) and 23% (TEE) of patients with acute aortic dissection. 7,3 Angiographic abnormalities detected by conventional aortography were observed in four of the 21 patients (traumatic and nontraumatic) enrolled in this study. The low rate of detection of AIH by aortography is explained by the impossibility to opacify the false channel because of the absence of an entrance tear. :4 The diagnosis of AIH by TEE relies on the detection of localized increased aortic wall thickness (>5 mm). 3 This segmental thickening of the aortic wall was circular or crescentic and had a thrombuslike echocardiographic appearance (Fig. 12). In patients with nontraumatic AIH, the aortic wall thickening was usually crescentic (14 of 15), whereas in patients with traumatic AIH the thickening was generally circular (5 of 6). Fifty-seven

American Heart journal Volume 134, Number 3

5 0 6 Vilacosta et ak

percent of our patients (12 of 21) (traumatic and nontraumatic) had echo-free spaces within the intramural hematoma probably representing liquid areas. When these echolucent areas were located immediately beneath the intimomedial flap, a

nonmobile dissection flap could be seen. In our series this flap was seen in four (19°/0) patients. Therefore the visualization of the dissection flap per se should not be

an exclusion criteria for diagnosing AIH. We do not know if these echo-free spaces are a sign of active bleeding, that is, vasa vasorum hemorrhage that has not yet clotted. If so, the lack of these echolucent areas could portend a better prognosis in terms of progression because this would suggest that bleeding has stopped. In this series the presence or absence of these areas had no prognostic implications. Because of the absence of an entrance tear the presence of blood flow within the intramural hematoma is an unfrequent finding. 3 In most patients, the false channel has not

been decompressed by reentry tear formation and, therefore, there is no choice for detecting color Doppler flow signals in these patients. Nonetheless, in a few patients there are some communicating points between the hematoma and the true lumen, probably caused by the pressure generated by the intramural hematoma. These points permitted the entrance of slow blood flow into the false channel. Four (19%) patients in this series had low-velocity color-Doppler flow signals within the echo-free areas of the intramural hematoma. Most importantly, a definitive echocardiographic aspect in the diagnosis of patients with AIH is the dynamic behavior of the hematoma. 3 The thickened aortic wall tends to vary with time. Spontaneous partial regression

(resorption) and even disappearance (normalization of wall thickening) of the AIH was seen in 10 patients (47%) in this series. Progression (increased thickening) of AIH was detected in one patient. Development of a classic dissection occurred in three patients (14%). One patient (patient 1) had early AIH regression and late intimal disruption (Fig. 13). It is worthy to emphasize that aortic insufficiency and pericardial effusion can occur in AIH without an associated dissection flap. Discriminating between AIH and other aortic lesions, including classic DeBakey type III aortic dissection with thrombosed false lumen and severe atherosclerosis with segmental wall thickening, may be clinically difficult and echocardiographically confusing. Our patients always had a second diagnostic imaging technique (CT or MRI) to confirm the echocardiographic diagnosis. The high density of fresh hematoma on CT scans is

specific for AIH, and MRI permits assessment of the age

of the hematoma on the basis of the high signal

intensity exhibited by the formation of methemoglobin.

At the initial examination some echocardiographic clues

will help to differentiate AIH from the aforementioned

aortic pathologies (Table III). The chronic descending

aortic dissection with a clotted false lumen frequently

exhibits the presence of a dissection flap, central

displacement of calcium, and echo-free spaces; in

addition, the type of thickening is crescentic. The

echocardiographic appearance of a severe

atherosclerosis in the descending aorta is commonly

depicted by a dilated aorta with spontaneous contrast

and absence of signs of dissection within the thickened

aortic wall (flap, flow, central displacement of calcium

and echo-free areas). The echocardiographic

characteristics of a descending AIH have already been

commented on, but if after a thorough clinical

echocardiographic evaluation uncertainty persists one

has the recourse to appeal to the dynamic nature of this

disease, and if echocardiographic follow-up changes are

detected the diagnosis of AIH is likely.

In this study we have shown that AIH can be present

in patients who are involved in violent deceleration

accidents who are seen with a widened mediastinal

silhouette on chest radiograph. Parmley et al., 15 on the

basis of anatomic specimens obtained from patients

with traumatic injuries, observed the presence of aortic

intimal hemorrhage. By TEE, intimomedial laceration of

the aorta, false aneurysm formation, and partial or

complete aortic disruption in patients with traumatic

aortic injuries have been described. ~6 Therefore our

observation of AIH in these patients is not surprising.

Usually trauma patients undergo aortography to role out

the presence of traumatic disruption of the aorta and, as

has been demonstrated in this study, this technique will

fail in detecting the existence of an AIH. All traumatic

AIH were located in the descending thoracic aorta in

the region of the aortic isthmus. This finding is in

agreement with the results of Vignon et al. 16 and Smith

et al. 17 In this traumatic series there was no patient in

whom the aortic injury involved the ascending aorta.

Mediastinal hematomas are relatively common in

trauma patients because of associated laceration of

small mediastinal vessels. Interestingly, the most

important differential echocardiographic sign between AIH and mediastinal hematoma is that in the former

there is a decrease of the aortic lumen diameter.

American Heart Journd September 1997

ViJacosta et al. 507

Prognosis of AIH Because different pathophysiologic processes lead to

intramural hemorrhage, the prognosis of nontraumatic and traumatic AIH is different. In this series the prognosis of patients with nontraumatic AIH confined to the descending aorta was good. Only one patient of seven died; he had adventitial rupture and required surgery. The outcome of patients with nontraumatic AIH that involved the ascending aorta is much worse. Two patients died suddenly, two resulted in a classic aortic dissection, and in one patient a progression of the AIH was documented. In contrast, patients with traumatic AIH had a good echocardiographic follow-up and their prognosis depended on the patient's situation. In this subgroup only one trauma patient of six resuIted in a classic dissection, two patients died from multiorgan failure, and the remaining three patients did well. In this series the degree of aortic wall thickening had no prognostic implications.

Conclusions First, AIH is an infrequent variant of aortic dissection

that can be detected and monitored by TEE but not by conventional aortography. Second, two types of MH can be distinguished: (1) traumatic of good prognosis and (2) nontraumatic, which can be an early stage of the classic aortic dissection with bad prognosis in cases involving the ascending aorta.

References 1. Roberts WC. Aortic dissection: anatomy, consequences, and causes.

Am Heartj 1981;101:195-214. 2. Vilacosta I, Castillo JA, San Ram6n JA, Roll6n M J, Aragoncillo P,

Sdnchez-Harguindey L. New echo-anatomical correlations in aortic dissection. Eur H cart J ] 995116:126-8.

3. Mohr-I<ahaly S, Erbel R, Kearny P, Puth M, MeyerJ. Aortic intramural hemorrhage visualized by transesophageal echocardiography: findings and prognostic implications, j Am Coil Cardio[ ]994123:658 64.

4. Krukenberg E. Beitr@ge zur frage des aneurysma dissecans. 8eitr Path Anat AIIg Path 1920;67:329-51.

5. Yamada T, Tada S, Harada J. Aortic dissection without intimal rupture: diagnosis with MR imaging and CT. Radiology 1988; ]68:347-52.

6. Robbins RC, McManus RP, Mitchell RS, Latter DR, Moon MR, Qlinger GN, et al. Management of patients with intramural hematoma of the thoracic aorta. Circulation 1993; 88 suppl Ih [I-1-10.

Z Nienaber CA, van Kodolitsch Y, Petersen B, Loose R, Helmchen U, Haverich A, el ah Intramural hemorrhage of the thoracic aorta. Diagnostic and therapeutic implications. Circulation 1995;92:1465-72.

8. Erbel R, Engberding R, Daniel W, Roelandt J, Visser C, Rennollet H. Echocardiagraphy in diagnosis of aortic dissection. Lancet 198911:457-60.

9. Vilacosta I, Castillo JA, Petal V, Batlle E, Roll6n M J, S6nchez Harguindey L. Intramural aortic hematoma following intraaortic balloon counterpulsation. Documentation by transesophageal echocardiography. Eur HeartJ 1995116:20154.

10. Hirst AE, Johns Vj, Kime SW. Dissecting aneurysm of the aorta: a review of 505 cases. Medicine 1958;37:217-79.

11. Lui RC, Menkis AH, Neil Mckenzie F. Aortic dissection without intimal rupture: diagnosis and management. Ann Thorac Surg 1992;53:886-8.

12. Stanson AW, Welch TJ, Ehman RL, Sheedy PF I]. A variant of aortic dissection: computer tomography and magnetic resonance findings. Cardiovasc Imaging 198911:55-9.

13. Vilacosta I, San Rom6n A, Peral V, CastilJo J, Dominguez L, BatHe E, et ak Imaging aortic intramural hematoma. Identification of two groups of patients [abstract]. Circulation 1995;92 suppl 1:1-307

14. Bansal RC, Chandrasekaran K, Ayala K, Smith DC. Frequency and explanation of false negative diagnosis of aortic dissection by aortography and transesophageal echocardiography. J Am Coil Cardiol 1995;25:1393-401.

15. Parmley LF, Mattingly TW, Manion WC, Jahnke Ej. Nonpenetrating traumatic injury of the aorta. Chculation 1958;17:1086-101.

16. Vignon P, Gu6ret P, Vedrinne j, Lagrange P, Cornu E, Abrieu O, et al. Role of transesophagea[ echocardiography in the diagnosis and management of traumatic aortic disruption. Circulation 1995;92:2959-68.

17 Smith MD, Cassidy JM, Souther S, Morris EJ, Sapin PM, Johnson SB, et ak Transesophageal echocardiography in the diagnosis of traumatic rupture of the aorta. N Engl J Med 1995;332:356-62.

American Heart Journal Vdume 134, Number 3