JACC Vol . 24, No . 2August 1994A94603

Intravescular Ult sound Inn

Objectives . This study was performed to determine the poten-tial of intravascular ultrasound in the detection and delineation ofaortic dissection .

Vackground. Intravascular ultrasound is a new techniquecapable of displaying real-time cross-sectional images of arterialvasculature. Its clinical use has been explored mostly in coronaryand peripheral arterial circulation .

bietkods. Intravascular ultrasound imaging of the aorta wasperformed using a 20-MHz ultrasound catheter in 28 patients withsuspected aortic dissection . All patients underwent contrast an-giography ; 7 had computed tomography and 22 had transesoph-ageal echocardlography.

Resalts. Imaging of the aorta from the root level to itsbifurcation was performed in all patients in an average of 10 min .No complications occurred . Dissection was present in 23 patientsand absent in 5 . In the patients without dissection, intravascular

Intravascular ultrasound is a newly emerging method capa-ble of providing high resolution cross-sectional images ofnormal and abnormal arterial architecture (1) . Currently, theprimary focus of work with this new imaging tool is directedtoward the evaluation of small arterial vasculature, such asthe coronary and peripheral vessels (2-11) . A catheter-basedapproach, this imaging method has the potential to displayeven larger vessels, such as pulmonary arteries and aorta .Although recent investigations have documented its possibleuse in the pulmonary circulation (12-15), its value in theevaluation of aortic disorders such as aortic dissection hasnot been well examined . Case reports and preliminary

From the New England Medical Center, Tufts University School ofMedicine, Boston, Massachusetts ; *The University of Mainz, Mainz, Ger-many, tUniversity of Bari, Bad, Italy ; and tUniversity Hospital, San Carlos,Madrid, Spain . This study was carried out with local institutional support .

Manuscript accepted March 11, 1993 ; revised manuscript received March4, 1994, accepted March 11, 1994 .

Address for corre%jaaduu : Dr. Natesa G. Pandian, Tufts-New EnglandMedical Center, Box 32, 750 Washington Street, Boston, Massachusetts02111 .

01994 by the American College of Cardiology

1A

NEW METHODS

te A011C Dissection

ANDREW R. WEINTRAUB, MD, FACC, RAIMUND ERBEL, MD, FACC,*GUNTER GORGE, MD,* STEVEN L. SCHWARTZ, MD, FACC, JUNBO GE, MD,*THOMAS GERBER, MD,* JURGEN MEYER, MD,* TSUI-LIEH HSU, MD,ROBERT BOJAR, MD, FACC, SABINO ILICETO, MD, FACCJ L . CARELLA, MDJPAOLO RIZZON, MIA, ISIDRE VILACOSTA, MD,f JAVIER GOICOLEA, MD,tJOSE ZAMORANO, MD,f FERNANDO ALFONSO, MD,f NATESA G . PANDIAN, MD, FACCBoston, Massachusetts; Mainz, Germany; Bari, Italy ; and Madrid, Spain

ultrasound revealed normal nortle anatomy . In all 23 patients withdissection, intravascular ultrasound demonstrated the intimal flapand true and false lumena . The longitudinal and circumferentialextent of aortic dissection, contents of the false lumen, involve-ment of branch vessels and 03e presence of intramural hematotnain the aortic wall could also be identified . In cases where aortog-raphy could not define the distal extent of the dissection, intra-vascular ultrasound did .

Conclusions. Our experience in this series of patients withaortic dissection indicates that intravascular ultrasound couldvaluable in the identification and categorization of aortic dissec-tion and in the description of associated pathologic changes thatmay be clinically important . It can be performed rapidly andsafely and could serve as an alternative or adjunct diagnosticprocedure in patients with aortic dissection .

(J Am Coll Car&ol 1994,24 .-495-303)

clinical experience suggest that intravascular ultrasoundcould be useful in delineating aortic coarctation and aorticatherosclerosis (16,17) . After experimental demonstrationthat intravascular ultrasound could detect dissections insmall and large vessels (18), we first reported the use ofintravascular ultrasound in a patient with acute aortic dis-section (19) . In this patient, images of the aorta by intravas-cular ultrasound crisply displayed the circumferential andlongitudinal extent of the dissection that were not clearlyevident by either contrast aortography or computed tomog-raphy. Findings in this and in two other subsequent casereports suggested that intravascular ultrasound could pro-vide a means for rapidly obtaining real-time, high resolutionimages of the aortic dissection and the associated pathology(19-21). Although methods such as transesophageal echo-cardiography, computed tomography and contrast aortogra-phy are helpful in this condition, there are limitations relatedto these methods (22-27). We report our observations on theintravascular ultrasound features in a series of patients withaortic dissection during an extended clinical experience withthis procedure .

0735-1097/94/$7 .00

495

496

MethodsTwenty-eight patients with suspected aortic dissection

were evaluated by intravascular ultrasound imaging . In 23 ofthese patients (12 men, 11 women ; 28 to 79 years old),clinical history and physical examination were compatiblewith • or strongly suggestive of acute aortic dissection . All 23had contrast aortography; 22 patients underwent transesoph-ageal echocardiography as well ; 7 had computed tomogra-phy. Nine patients underwent surgery . Intravascular ultra-sound imaging was performed after either aortography,transesophageal echocardiography or computed tomogra-phy . Intravascular ultrasound imaging was performed in fiveother patients who had vague chest, shoulder and back painsymptoms but with normal coronary arteriography ; how-ever, the probability of aortic dissection, although small,needed to be excluded in these patients . Intravascular ultra-sound imaging was performed as the first diagnostic proce-dure in this group, followed by aortography . Informedwritten consent was obtained from all patients before theprocedure, which was performed in accordance with theapproval of the individual human investigation committees .

Intravascular ultrasound imaging was performed with acommercially available instrument (2,12) . This device in-cludes a 6F or 4.8F, 95- or 100-cm long disposable catheterenclosing a mechanically rotating driveshaft with a 20-MHzultrasound crystal at its tip (Sonicath, Boston ScientificCorporation). The catheters were interfaced with an imagingconsole adapted for 20-MHz operation and 360 0 scans (Dia-sonics). The ultrasound core is rotated by a motor at 900 rpm(iS framesls), delivering a real-time, circumferential ultra-sound image. The catheter has a proximal tract at the distalend of the catheter and a distal tract at the catheter tip thataccept a 0.025-in. (0.635 cm) guide wire . The ultrasoundcatheter Was introduced by way of an 8F femoral arterysheath and was advanced over the guide wire under fluoro-scopic guidance . Images were acquired at various levels ofthe proximal aorta, including aortic root, aortic arch, de-scending thoracic aorta and abdominal aorta. Adjustments ofcatheter position, ultrasound gain, depth gain compensationand gray scale were made to obtain optimal images (28). Allstudies were recorded on a videotape for subsequent analy-sis .

Dab analysiL Each diagnostic test (intravascular ultra-sound, transesophageal echocardiography, aortic angiogra .phy, and computed tomography) was interpreted in anunblinded fashion . For each imaging modality, the extent ofdissection was defined using the DeBakey classificationsystem (29) .

criteda for Imaging techniques. Contrast an-giography. Aortic dissection was defined by visualizationof a double lumen or intimal flap (25,30). The location of theintimal tear was identified by a discrete site in which contrastwas seen to pass from one lumen to the other . Aorticregurgitation was diagnosed when contrast entered the leftventricle .

),J

14

WEINTRAUB ET AL.INTRAVASCULAR ULTRASOUND IN AORTIC DISSECTION

JACC Vol . 24. No . 2August 1994:495-503

Computed tomography . Visualization of two distinct lu-mens separated by an intirrlal flap or displacement of intimalcalcium was required for the diagnosis of aortic dissection(23,24) . The presence of a low attenuation crescentic regionalong the aortic wall was identified as aortic wall hematoma(31) .

Transesophageal echocardiography. The presence of anintimal flap separating a true and false lumen was evidencefor the diagnosis of dissection (22) . Aortic wall hematomawas defined as a focal, crescent-shaped thickening of theaortic wall, usually containing a hypoechoic area (32) . Entrysites were identified when there was disruption in the conti-nuity of the flap or a discrete flow jet toward the false lumenthat emanated from a well defined location of the intimal flap(33) . Thrombus was defined as a collection of soft, granularechoes within the false lumen (34) . Aortic insufficiency wasdiagnosed when color flow or spectral Doppler detecteddiastolic turbulent flow into left ventricular outflow tractfrom the aorta.

Inaravasculaar ultrasound . Aortic dissection was deemedpresent when an intimal flap, a curvilinear structure withrapid, often chaotic motion, could be identified separatingthe aorta into a true and false lumen (19-21) . As withtransesophageal echocardiography, thrombus was diagnosedby the appearance of soft granular echo signals and aorticwall hematoma when a focal area of thickening occurred thatappeared atypical for plaque . Sites of perforation in theintimal flap were identified as gaps in the flap with asynchro-nous motion on either side of the tear. This was distin-guished from "dropout ." a localized area with loss ofsignals .

Static analysis. Sensitivity and specificity were calcu-lated by comparing the results obtained with intravascularultrasound for the diagnosis of dissection, branch involve-ment with angiography and thrombus or hematoma in thefalse lumen with transesophageal echocardiography. The95 confidence intervals were calculated using the normalapproximation to binomial distribution .

Results

Intravascular ultrasound examination in the 28 patientsdemonstrated aortic dissection in 23 patients and a normalaorta in 5 . In the 23 patients with clinically suspected acuteaortic dissection, aortography and another technique, trans-esophageal echocardiography or computed tomography,demonstrated evidence of aortic dissection . Comparativeresults of each imaging modality are summarized in Table 1 .Eleven patients had DeBakey type I dissection, five had type11, and seven had type 111 . In each of these patients,intravascular ultrasound imaging clearly identified the pres-ence and extent of aortic dissection . In the five patients whohad intravascular ultrasound as the primary procedure toexclude aortic dissection, no evidence of dissection wasnoted on the ultrasound images (Fig . 1). Aortography inthese patients was normal . The sensitivity and specifity for

0 = 211

intravascular ultrasound in the diagnosis of aortic dissectionwas 100, The average time required for intravascularultrasound imaging was 10 min. There were no complica-tions related to the imaging procedure .

Intravascular ultrasound findings for aortic dissection .Whereas normal aorta was seen as a circular cross-sectionalimage with an intact aortic wall and a clear lumen, aorticdissection was displayed on the two-dimensional intravascu-lar ultrasound images as disruption of the aortic wall, withthe intimal layer prqjecting into the aortic lumen, separatingtrue and false channels (Fig . 1 to 3). The appearance wassimilar to that seen on transesophaged echocardiograms,except that the imaging field was circular and the ultrasoundcatheter and the guide wire were seen within the true lumen(Fig. 4). The intimal flap could be seen either aligned withone area of the wall if the dissection was small or coursingacross the lumen if the dissection was extensive . In all cases,the curvature, shape and mobility of the intimal flap wereseen . Locations where the intima was separated from thewall and those where it remained attached could be welldefined, allowing us to assess the circumferential extent ofdissection. The pulsatile motion of the intimal flap duringsystole was helpful in the differentiation of the true and falselumen. In some patients, it was difficult to obtain completecross-sectional images of the vessel within a single frame ofthe image display at the aortic arch level and at locationswhere the aorta was aneurysmally large, because of thedifficulty in maintaining the ultrasound catheter in a centraland coaxial orientation and because of the limited depth offield associated with 20-MHz catheters . Despite this prob-

Figure 1 . Intravascular ultrasoundimages from a patient with a nor-mal descending thoracic aorta(left) and a patient with DeBakeytype III aortic dissection (right) .The ultrasound catheter (c) isnoted within the true lumen (M) .The aortic wall appears intact inthe normal subject (open arrows) ;the intimal layer (small solid arrow)appears separated from the wall inthe patient with dissection . Boththe true and false lumena (L) areseen as relatively sonolucent areas .

lens, we were able to verify whether the arch was involved inthe dissection because it was possible to interrogate mostparts of the arch and other enlarged regions of the aorta insome orientation, even when perfect cross-sectional orien-tation was not available (Fig . 5 and 6). Precise delineation ofthe entry site tears in the intima was possible in six patients(Fig. 6) . However, recognition of the beginning of thedissection in other patients was not difficult because theseparated intima could be visualized in all, if not the entrytear . In all 23 patients, the proximal site of intimal separationand the distal end oC the dissection could be recognized, andthe categorization of the type of aortic dissection by intra-vascular ultrasound, whether type 1, II or III, correspondedto the observations by aortography, transesophageal echo-cardiography or computed tomography .

The true and false lumena and their relative cross-sectional areas at different levels could be discerned byintravascular ultrasound in all patients . Figures l to 4 showexamples displaying various degrees of lumen compromise .Only a moderate reduction in lumen area is noted in theexample shown in Figure 1 ; Figure 4 demonstrates a truelumen that accounts for only a very small portion of theoverall aortic cross-sectional area. Real-time imaging alsorevealed the spiral nature of the intimal tear often encoun-tered in aortic dissection . In some cases, the fast-flowingblood in the true lumen and relatively slow-flowing blood inthe false lumen exhibited granular moving signals that wererecognizable by intravascular ultrasound .

Changes in the aortic wall due to hemorrhage into themedia were also visualized by intravascular ultrasound . In

JACC Vol . 24 . No. 2August 1994:495-503

WEINTRAUS ET AL.iNTRAVASCULAR ULTRASOUND IN AORTIC DISSECTION

497

Table 1. Comparison of Intravascular Ultrasound With Other Imaging Modalities in Patients With Suspected Aortic Dissection

DissectionEntry SiteDefined

CircumferentialExtent

LongitudinalExtent

BranchInvolvement

Thrombus/Hematoma

Intravascular ultrasound (n = 28) 23 6 23 23 8 4Anytogram (n = 29) 23 15 0 20 9 0Computed tomography (n - 8) 8 7 8 8 0 1Transesophageal echo cardiography 21 18 21 5 0 8

498 WEINTRAUB ET AL .

JACC Vmt, 24, No. 2INTRAVASCULAR ULTRASOUND IN AORTIC DISSECTION

August 1994 : 495 -503

F' 2. Intravascular ultrasound im-ages from a patient with aortic dissec-tion . The schematic (middle) depicts thelevels at which the images were ac-quired. The intimal flap and the true (TL)and false (FL) lumena are well visual-ized . The intimal flap is seen to extendinto the right iliac artery (arrow) . cultrasound catheter .

Figure 3. Aortogram (top left) and fluoroscopic(top right) and intravascular ultrasound (bot .tom) images front a patient with type I aorticdissection . Aortogram shows an extensive dis-section flap ( w) coursing through the aortictree . The fluoroscopic image demu istrates theultrasound catheter tip (arrow) in the descend-ing thoracic aorta . A guide wire is noted extend-ing to the level of the aortic arch, and a pul-monary artery catheter is also seen . In theintravascular ultrasound image, the ultrasoundcatheter is seen as a small ring (C) . The truelumen (TL) of the aorta is separated from thefalse lumen (FL) by the dissection flap . Thesurrounding adventitia of the aorta is seen asthe outermost hyperechoic structure .

JACC Vol . 24, No . 2August 094 :495-503

Figure 4 . Transesophageal echo-cardiographic (left) and iniravas-cular ultrasound (riWbt) imagesfrom a patient with a type I dis-section . Both techniques displaythe presence of extensive dissec-tion and marked compromise ofthe true lumen MY FL = falselumen .

one patient who was found at surgery to have a largedissection flap and extravasation of the blood through theadventitia in the thoracic aorta, transo-ophageal echocardi-ography revealed extensive dissection of the aortic wall,with multiple dissection tracts within the wall of the aorta .These findings were also noted by intravascular ultrasound(Fig . 7) . The intramural hematoma was displayed on intra-vascular ultrasound as increased wall thickness and anirregular appearance of the wall architecture . In this patient,the dissection and mural abnormalities extended to the levelof the lower abdominal aorta by catheter ultrasound .

In all patients, the distal end of aortic dissection could bedefined by intravascular ultrasound . In three patients, aorticdissection was seen to extend distally to the iliofemoralvessels (Fig . 2). The involvement, or lack thereof, of otheraortic branches could be recognized in 14 patients . Overall,the sensitivity for intravascular ultrasound to detect involve-

Figure S. Intravascular ultrasound image at the level of the aorticarch from a patient with type I dissection . The dissection involvesthe arch (left) with dissection flap (arrows) separating the false (FL)from the true lumen . The innominate artery (right) is seen originat-ing (arrows) from the aorta (AO) . The dissection flap is shown toinvolve this branch (arrowheads) with a small false lumen (asterisk) .

WEENTRAUB ET AL .

499INTRAVASCULAR ULTRASOUND IN AORTIC DISSECTION

ment of a branch vessel in the dissection was 89% (95%confidence interval [CI] 78% to 100%) ; specificity was 100% .Figure 8 is an intravascular ultrasound image from a patientwith type III aortic dissection . In this patient, intravascularultrasound revealed that the celiac artery appeared to origi-nate from the true lumen. At the renal artery level intravas-cular ultrasound demonstrated that the left renal arterycommunicated with the false lumen, whereas the right renalbranch was in continuity with the true lumen . This wasconsistent with the findings on aortography .

Comparison of intravascular ultrasound with transesopha-geal echocardiography, computed tomography and aortogrn-phy. Twenty-one patients had transesophageal echocardiog-raphy and intravascular ultrasound imaging (Table 1) . Theabnormalities noted in terms of the presence and circumfer-ential extent of aortic dissection and the lumen characteris-tics were similar with both techniques (Fig . 4 and 7) . In sixpatients who had a dissection that involved the abdominalaorta, transesophageal echocardiography was unable todemonstrate the distal extent ; conversely, intravascular ul-trasound clearly defined the distal-most extent of aorticdissection . In three of these patients, the iliac or renal artery

Figure 6. Intravascular ultrasound image at the level of the aorticarch from a patient with a type I dissection . An entry tear (arrow) inthe intima is seen . In real time, the edges of the intima adjacent tothe tear demonstrated chaotic motion . Abbreviations as in Figure I -

500 WEINTRAUB ET AL .INTRAVASCULAR ULTRASOUND IN AORTIC DISSECTION

Figure 7, Transesophageal echocardiographic (top) and intravascu-tar ultrasound (bottom) images of the descending thoracic aorta froma patient with a type III dissection. The aortic walll in this region isthiclrened by the presence of hematoma within the aortic wall,which was confirmed at operation. The extent of intimal disruptionand th intramural hematoma ( )have a similar appearance inboth images. Abbreviations as in Figure l .

was affected. Understandably, this was not demonstrable bytransesophageal echocardiography, although intravascularultrasound identified the branch involvement . Dissection of

JACC Vol. 24, No . 2August 1994 :495-503

the ascending aorta could be identified by both techniques ;however, transesophageal echocardiography depicted thedissection in this part of the aorta more clearly than intra-vascular ultrasound . Multiple communications between thetrue and false lumena often identifiable by transesophagealcolor Doppler were not demonstrated by intravascular ultra-sound. Transesophageal echocardiography also denoted thelack of involvement of coronary arteries in nine patients withtype I or Il dissection, but intravascular ultrasound could notprovide that information. Eight patients had thrombus oraortic wall hematoma evident on the transesophagealechocardiogram; these findings were noted on intravascularultrasound in only four patients. Compared with transesoph-ageal echocardiography, the sensitivity of intravascular ul-trasound in the detection of thrombus or hematoma in thefalse lumen was 50°1 (95% Cl 15% to 85%) ; specificity was100,70 .

Eight patients underwent computed tomographic andintravascular ultrasound imaging . Computed tomographydid not provide any important additional information ; in fact,in four patients computed tomography was unable to definethe distal end of the dissection or other dissection-relatedabnormalities as clearly as intravascular ultrasound . Al-though the aortic wall was seen by both techniques, intra-vascular ultrasound provided higher resolution images of theaortic wall and intimal flap . However, the aorta was seen inits entire cross section irrespective of its size by computedtomography at all levels, whereas the 20-MHz ultrasoundcatheters we used could not display the whole aorta at thearch level and at regions of aneurysmal dilation .

All patients had contrast aortography, and in all bothintravascular ultrasound and contrast aortography detecteddissection . Aortic regurgitation could be seen by aortogra-phy in three patients but not by intravascular ultrasound .Extension of intimal tear into the innominate artery wasevident in one patient by aortography but not by intravas-cular ultrasound . In another patient, continuation of thedissection into the right iliac artery was not shown byaortography but was observed by intravascular ultrasound .The distal portion of dissection in the abdominal aorta wasbetter defined by intravascular ultrasound than aortographyin three other patients.

8. intravascular ultrasound images of the ab-dominal aorta in a patient with type III dissectionobtained at the level of the celiac artery (left) and at thelevel of the renal arteries (right) . In the image on theleft, the celiac artery appears to arise from the truelumen of the aorta (arrow). The ultrasound catheterring signal is seen in the true lumen . The dissectionflap is seen separating the true from the false lumen. Inthe image at the level of the renal arteries, the rightrenal artery is in communication with the true lumen(solid arrow) ; the left renal artery is supplied by thefalse lumen (open arrow) . These relations were betteridentified in the real-time images .

JACC Vol. 24, No . 2August 1994 :495-503

DMadonOur experience with the use of intravascular ultra-sound

imaging in this series of patients with aortic dissectionindicates that this method of imaging could be valuable in theidentification and categorization of aortic dissection . Inaddition, catheter-based ultrasound appears to have thecapability to describe many of the pathologic changes relatedto aortic dissection, such as degree of lumen compromise,presence or absence of c!ol-: in the false lumen and involve-ment of aortic 'ranches . Our observations also point to therelative ease and safety of this procedure .

Detection and categorization of aortic dissection . Intravas-cular ultrasound identified aortic dissection equally as wellas transesophageal echocardiography, computed tomog-raphy and aortography. Similarly, it was able to excludedissection as reliably as contrast aortography . Delineation ofthe type of aortic dissection was also possible because theaorta could be imaged in its entirety from the level of theaortic root ; thus the proximal and distal points of intimal tearcould thus be identified . Findings in this study indicate thatintravascular ultrasound is probably superior to other tech-niques in defining the distal extent of aortic involvement .The problem of a catheter entering the false lumen is wellknown in the performance of aortography . Initial ultrasoundexamination could determine whether the catheter is in thetrue or false lumen before contrast angiography, therebypreventing inadvertent injection into the false lumen andextension of the dissection . The ability to visualize thevessel lumen from the moment of introducing the ultrasoundcatheter into the femoral artery makes it possible to firstrecognize the distal end of the dissection process by theimages of the intimal flap . As one enters the region of aorticdissection, the curvature and pulsatile dynamics of theintimal flap aid in ensuring that the ultrasound catheter isindeed in the true lumen. During intravascular ultrasoundexamination, as the catheter is advanced in a retrogrademanner into the more proximal portions of the aorta withcontinuous imaging, the torn aortic wall and other relatedabnormalities become evident . If the cross-sectional size ofthe aorta exceeds the imaging field, periodic reorientationof the ultrasound catheter allows examination of all quad-rants of the aortic wall at any level. It is not surprising,therefore, that detection of aortic dissection and determina-tion of its extent were possible by intravascular ultrasound inevery case in this series . Although our experience in thisstudy suggests that intravascular ultrasound is highly reliablein recognizing the presence or absence of aortic dissection,inferences as to diagnostic sensitivity and specificity are notpossible because the series is small, the patients includedwere not consecutive, and the study was not performedprospectively with blinded interpretation .

Depiction of abnormalities associated with aortic dissec-tion . In addition to determining the type of aortic dissection,intravascular ultrasound appears to be capable of providinginformation on other aspects of this disorder that may be

WEINTPAUB ET AL .INTRAVASCULAR ULTRASOUND IN AORTIC DISSECTION

501

essential for optimal patient management (27) . The degree oflumen compromise, presence or absence of thrombus in thefalse lumen, intramural hematoma, periaortic extravascularhematoma and branch involvement can be ascertained .Similar to transesophageal echocardiography, intravascularultrasound depicts the intimal flap and the true and falsechannels, and the circumferential extent of the tear is easilydefined . Unlike transesophageal echocardiography, intra-vascular ultrasound clearly displays the dissection process inthe abdominal aorta and the involvement of the abdominalaortic branches . Knowledge about branch involveniunt-dissection into the renal arteries, for example-may influ-ence therapeutic strategies in type III dissection, for which aconservative medical management is usually employed . De-tection of the involvement of aortic arch branches andcoronary arteries was difficult with 20-MHz devices but mayhe easier with lower frequency ultrasound catheters . Throm-bosis of the false lumen may influence management deci-sions in some patients (35), and the intraluminal clot isdisplayed by intravascular ultrasound . Intramural dissec-tions and extravascular clots may indicate rupture or im-pending rupture (32) ; our observations suggest that thiscomplication may be identified by intravascular ultrasound .Pericardial effusion, another complication of dissection, wasnot observed with intravascular ultrasound in this study . Wehave previously shown, however, that effusions can bereadily detected by placing the transducer within the heartitself (36,37) . If information concerning ventricular functionor effusion is desired, it can be potentially provided bycontinued advancement of the catheter across the aorticvalve into the left ventricle . At the current state of intravas-cular ultrasound technology, detection of aortic insufficiencyis not possible by this method . Howe"er, this may not be amajor drawback because most patients with aortic dissectionoften undergo conventional transthoracic echocardiography,which readily displays aortic regurgitation even when it isunable to depict the dissection .

Comparison of intravascular ultrasound with other imag-ing modalities. It cannot be stated that intravascular ultra-sound is superior to other diagnostic techniques used indetecting aortic dissection, such as transesophageal echoca --diography, contrast aortography, computed tomography andmagnetic resonance imaging . On the basis of our experience,however, we suggest that this imaging modality appears tohave the potential to provide similar, if not identical, infor-mation. Transesophageal echocardiography is probably themost useful method to detect and define aortic dissectioneven though it cannot depict the abdominal aorta (22,38,39) .Although intravascular ultrasound may not provide all thedata supplied by transesophageal echocardiography, theamount of in-formation available by intravascular ultrasound(proximal site, type, extent, lumen contents, mural abnor-malities) may be adequate for patient management in manycases. Computed tomography and magnetic resonance im-aging are useful in aortic dissection, but it takes considerabletime to mobilize the team and acquire data. Contrast aortog-

5112 WEINTRAUB ET AL .INTRAVASCULAR ULTRASOUND IN AORTIC DISSECTION

raphy defines the branch involvement and aortic regurgita-tion well, but this technique has well known limitations : Itrequires radio-opaque contrast medium ; it often does notclearly define the intimal flap and extent of dissection(particularly in the distal end) ; the wall of the aorta is notvisualized; and false negative diagnoses do occur, necessi-tating another test (26) . Intravascular ultrasound can beperformed without the need for radioopaque dye and can beconsidered a form of aortography without contrast medium,or it may serve as an adjunct to aortography, providingadditional information . If intravascular ultrasound is used inconjunction with sort phy, imaging first with this tecea-nique could ensure the placement of the aortography cathe-ter in the true rather than the false lumen, as occasionallyhappens. The major advantage of intravascular ultrasound isits capability to visualize the abdominal aorta and its sidebranches .

L future diredious. The major limitationsof our study were the small number of patients involved andthe fact that ultrasound studies were not subjected to blindedinterpretation. Although the sensitivity and specificity ofintravascular ultrasound were both 111 ",, and no complica-tions were observed, these data should be interpreted cau-tiously in light of the small number of patients and theunblinded review . The wide 95% confidence intervals fordiagnosis of branch involvement and presence of thrombusor hematoma reflect the small sample size . Some technicallimitations to intravascular ultrasound must also be consid-ered. Intravascular ultrasound is an invasive technique ;however, in our experience it was safely and easily per-formed, Experience with larger numbers of patients is re-quired to verify the safety of this technique .

Another major limitation pertains to the high frequency(20 MHz) ultrasound catheters used . Although 20-MHzcatheters provide excellent resolution images in small ves-sels, such as coronary and peripheral arteries, the depth ofimaging field is limited . This precludes complete visualiza-tion of the cross-sectional anatomy when the aorta is dilated .In half of the patients studied, the overall cross-sectionalarea of the aorta was too large to visualize in one imagingfield . Difficulties in aligning the catheter coaxially within the

may also result in only portions of the vessel imaged inone field, These problems occurred most commonly in theregion of the ascending aorta and aortic arch ; however, thisdid not affect the ability to diagnose dissection because anintilnal flap was visible in all regions. Although this may notaffect the ability to detect aortic dissection, the abnormali-ties associated with dissection, such as side branch involve-ment (as stated previously), may not be well defined in somepatients. This may be overcome by using devices capable ofi m ing to greater depths. Recently, ultrasound catheterswith lower frequency transducers (10, 12 .5 and 15 MHz)have been shown to allow an expanded depth of field and todisplay the great vessels and cardiac structures (37,40) . Suchproficiency may enhance the diagnostic capability of intra-vascular ultrasound in aortic dissection. However, because

JACC Vol. 24, No 2August 1994:495--

the current generation of intravascular ultrasound transduc-ers do not have Doppler capability, flow dynamics within theaorta and the presence or absence of aortic insufficiencycould not be assessed . Facilities for immediate three-dimensional reconstruction of intravascular ultrasound im-ages may increase the potential of this technique in aorticdissection (11,41). Intravascular ultrasound may also beiseful in new approaches being explored, such as endolumi-nal stenting for management of aortic dissection (42) . Fur-ther investigations using low frequency devices in largerseries of patients could define the precise clinical role ofintravascular ultrasound in aortic disorders .

Conclusions. Intravascular ultrasound is a potentiallyuseful tool for the diagnosis of aortic dissection and delin-eatson of the extent of dissection as well as branch vesselinvolvement. Although this technique is invasive, contrastagents are not required, and it can be performed separately,with or without concomitant contrast tangiography. Clinicaluses for this technique include diagnosis of dissection whentransesophageal echocardiography is contraindicated or un-desirable, peson s because of esophageal pathology or re-cent ingestion of a meal, or in addition to transesophagealechocardiography when further delineation of the dissectioninto the abdominal aorta is desired . Further studies withlarger numbers of patients are required before this promisingtechnique can be adopted for routine clinical use .

We thank Mrs . Linda Wing for valuable assistance in the preparation of themanuscript and William Rand. PhD, for assistance with the statistical analy-sis .

References

I. Pandian NO . Kreis A, Brockway 13, et at . Ultrasound ansioscopy :real-time, two-dimensional, intratursinal ultrasound imaging of bloodvessels. Am J Cardiol 1988 ;62:493-4 .

2. Pandian NO, Kreis A, Weintraub A, et at . Realtime intravascular ultra-sound imaging in humans. Am J Cardiol 1990;65 :1392-6 .

3. York PG, Johnson EL, Linker DT . intravascular ultrasound: develop-ment and clinical potential . Am J Cardiac lmagirg 1988 ;2 :185-93 .

4. Tobis JM. Mallery J, Mahon P, et al. intravascular ultrasound imaging ofhuman coronary arteries in vivo : analysis of tissue characterizations withcomparison to in vitro histological specimens. Circulation 1991 ;83 :913-26.

5 . Hodgeon JM, Graham SP. Savakus AD. et al . Clinical percutaneousimaging of coronary anatomy using an over-the-wire ultrasound cathetersystem . Int J Card Imaging 1989 ;4:187 43.

6. Potkin 8 .4, Bartorelli AL, Gessert 3M, et al . Coronary artery imagingwith intravascular high-frequency ultrasound. Circulation 1990;8l :1575-85 .

7. Nissen SE, Gripes CL, Gurley JC, et al . Application or a new phased-array ultrasound imaging catheter in the assessment of vascular dimen-sions . In vivo comparison to cineangiography . Circulation 1990;81:660-6 .

8 . Siegel RJ, Ariani M, Fishbein MC, et al . Histopathologic validation ofangioscopy and intravascular ultrasound . Circulation 1991 ;84 :109-17 .

9. Nishineura RA, Edwards WD, Warnes CA, et al . Intravascular ultrasoundimaging : in vitro validation and pathologic correlation . J Am Coll Cardiol1990 ;16:145-54.

10. Gussenhoven El, Essed CE, Lancee CT, et al . Arterial wall characteris-tics determined by intravascular ultrasound imaging : an in vitro study .I Am Coll Cardiol 1989 ;14:947-52.

11 . Rosenfield K, Losordo DW, Ramaswamy K, et al . Three-dimensional

JACC Vol. 24, No . 2August 1994 :495-503

reconstruction of i,irnan coronary and peripheral arteries from imagesrecorded during two-dimensional intravascular ultrasound examination .Circulation 1991 -,84:1938-56 .

12 . Pandian NO. Weintraub A. Kreis A, Schwartz SL, Konstam MA, SalemDN. Intracardiac, intravascular, two-dimensional, high-frequency ultra-sound imaging of r-Amonary artery and its branches in humans andanimals . Circulation 1990 :81 :2007-12 .

13 . Porter T, MoN.aty PK, Taylor D, Nixon JV, Pandian N . Loss of regionalvariations in pulmonary vascular reactivity in heart failure patients:evidence from simultaneous in vivo intravascular ultrasound evaluation ofpulmonary arteries in different lung fields [abstrartl . Circulation 1991 ;84Suppl 11 :11-594.

14 . Ricou F, Nicod PH, Moser KM, Peterson KL . Catheter-based intravas-cular ultrasound imaging of chronic thromboembolic pulmonary disease .Am J Cardiol 1991-,67 :749-52.

15 . Ricou F, Ludomirsky A, Weintraub RG, Saint DJ . Applications ofintravascular scanning and transesophageal echocardiography in congen-ital heart disease . lot J Card Imaging 1991 ;6 :221-30 .

16 . Harrison KJ, Sheikh I'EI, Davidson CJ, et al . Balloon angioplasty ofcoarctation of the aorta evaluated with intravascular ultrasound imaging .J Am Coll Cardiol 1991 ;t5:906-9.

17 . Weintraub A, Sanzobrino B, Smith 1, et al . Intravascular ultrasoundevaluation of aortic atherosclerosis and related disorders : aortographywithout contrast [abstract] . J Am Soc Echocardiogr 1991 ;4:291 .

18 . Pandian NO, Kreis A. Brockway B, Sacharoff A, Caro R . intravascularhigh-frequency two-dimensional ultrasound detection of arterial dissec-tion and intimal flaps . Am J Cardiol 1990,65 :1278-80 .

19 . Weintraub AR, Schwartz SL, Pandian NO, et al. Evaluation of acuteaortic dissection by intravascular ultrasonography [letter] . N Engl J Med1990;323 :1566-7 .

20 . Pande A, Meier B, Fieisch M, Kammerlander R, Siminet F, Lerch R .Intravascular ultrasound for diagnosis of aortic dissection . Am J Cardiol1991 -.67:662-3 .

21. Cavaye DM, French WJ, White RA, et al . Intravascular ultrasoundimaging of an acute dissecting aortic aneurysm : a case report . J Vase Surg1991,13 :510-7

22. Erbel R, Borner N . Stellar D, et al . Detection of aortic dissection bytransesophageal echocardiography . Br Heart J 1987 ;58 :45-51 .

23 . Hedberg E, Wolverson M . Sundaram M, Connors J, Susman N . CT scanfindings in thoracic aortic dissection . Am J Roentgenol 1981 ;136:13-20.

24 . White RD, Lipton MJ, Higgins C, et al. Noninvasive evaluation ofsuspected thoracic aortic disease by contrast enhanced computed tomog-raphy . Am J Cardiol 1986 ;57 :282-90 .

25 . Dinsmore RE, Rourke JA, DeSanctis RD, Harthorne JW, Austen WG .Angiographic findings in dissecting aortic aneurysm . N Engl J Med1966;275:1152-7.

26 . Shuford WH, Sybers RG, Weens HS. Problems in the aortographicdiagnosis of dissections of the thoracic aorta. N Engl J Med 1969,280 :225.31 .

27 . Treasure T, Raphael MJ . Investigation of suspected dissection of thethoracic aorta . Lancet 1991 :338 :490-5 .

WEINTRAUB ET AL.

503INTRAVASCULAR ULTRASOUND IN AORTIC DISSECTION

28. Weintraub A . Pandian NO . Impact of technical factors, catheter orienta-tion and vascular pathology on image acquisition, display and quantitativeanalysis of atherosclerotic arteries by high frequency intravascular ultra-sound [abstract]. J Am Soc Echocardiogr 1990;3 :214.

29. DeBakey M, Henry W, Cooley D, Morris G, Crawford E, Beall A .Surgical management of dissecting aneurysris of the aorta . J ThoracCardiovase Surg 1965 ;49 :130-49 .

30. Hayachi K, Meaney TF, Zelch JV, Tarar R . Aorlographic analysis ofaortic dissection . Am J Roentgenol 1974 ;122 :769-82 .

31 . Yamada T, Tada S, Harada J . Aortic dissection without intimal rupturediagnosis with MR imaging and CT. Radiology 19WI4357-68.

32. Zotz RI, Eke R, Meyer J. Noncommunicating intramural hematoma : anindication of developing aortic dissection? J Am Soc Echocardiogr1991 ;4:636-8 .

33. Hashimoto S, Kumada 1, Osakad G, et al . Assessment of transesopha-geal Doppler Echography in dissecting aortic aneurysm . J Am CollCardiol 1939;14:1253-62.

34. Pandian NQ Kreis A, Brockway B . Detection of intraarterial thrombusby intravascular high-frequency ultrasound : in vitro and in vivo studies .Am J Cardiol 1990 ;65 :1280-3,

35 . Erbel R, Oaten 11, Meyei J, et al . Effect of medical and surgical therapyon aortic dissection evaluated by transesophageal echocardiography :implications for prognosis and therapy . Circulation 1993X :1604-15 .

36. Weintraub AR, Schwartz SL, Smith J, Hsu TL ., Pandian NO . Intracardiactwo-dimensional echocardiography in patients with pericardial effusionand cardiac tamponade . J Am Soc Echocardiogr 1991 -,4:571-6 .

37. Schwartz SL, Gillam LD, Weintraub AR, et al . Intracardiac echocardi-ography in humans using a small sized (6 Fr), low frequency (12 .5 MHz)ultrasound catheter . Methods, imaging planes, and clinical experience .J Am Coll Cardiol 1993 ;21 :189-98.

38 . Erbel R, Enberding R, Daniel W, Roelandt J, Visser C, Rennoliet H .Echocardiography in the diagnosis of aortic dissection . Lancet 1989;1 :457-61 .

39. Cigarroa JE, Isselbacher EM, DeSanctis RW, Eagle KA . Diagnosticimaging in the evaluation of suspected aoric dissection : old standards andnew directions . N Engl J Med 1993 ;328 :35-43 .

43 . Pandian NG, Kumar R, Katz SE, et al . Real-time, intracardiac, two-dimensional echocardiography. Enhanced depth of field with a low-frequency (12 .5 MHz) ultrasound catheter . Echocardiography 1991 ;8 :407-22 .

41 . Chandrasekaran K, Sehgal CM, Hsu TL, et al . Three-dimensional intra-vascular ultrasound imaging of arterial atherosclerosis nd its complica-tions : improved recognition of the atheroma bulk, the span of dissectionand intimal flaps and the thrombus extent [abstract] . J Am Coll Cardiol1991 ;17 :233A .

42 . Cavaye DM, White RA, Lerman RD, et al . Usefulness of intravascularultrasound imaging for detecting experimentally induced aortic dissectionin dogs and for determining the effectiveness of endoluminal stenting . AmJ Cardiol 1992 ;69 :705-7.