Intracoronary ultrasound evaluation of interventional technologies

lntracoronary Ultrasound Evaluation of Interventional Technologies Charles J. Davidson, MD, Khalid H. Sheikh, MD, Katherine B. Kisslo, RDMS, Harry R. Phillips, MD, Robert H. Peter, MD, Victor S. Behar, MD, Yihong Kong, MD, Mitchell Krucoff, MD, E. Magnus Ohman, MD, James E. Tcheng, MD, and Richard S. Stack, MD

The feasibility and applicability of intravascular ultrasound (IVUS) of the coronary arteries were evaluated in 65 patients undergoing 70 coronary interventional procedures. Morphologic and quantitative analyses were performed with a mechanically rotated IVUS catheter (4.8Fr, 20 MHz) and with orthogonal view cineangiography. A semiautomated edge-detection algorithm was used for cineangiographic quantification. Coronary interventions included 45 percutaneous transluminal coronary angioplasties, 9 excimer lasers, 11 directional coronary atherectomies, 3 rotational atherectomies and 2 stents. Most lesions consisted of a mixture of plaque composition (hard, n = 30; soft, n = 64). Other unique morphologic data by IVUS were plaque topography (eccentric, n = 34; concentric, n = 36) and vessel dissection (IVUS [n = 291 versus angiography [n = 141, p
onventional contrast cineangiography provides a limited definition of vessel morphology at least partially due to silhouette imaging. However, cineangiography has been considered to be the gold standard for human in vivo assessmentof luminal dimensionsand extent of plaque disruption after coronary interventions. Previous studies have demonstrated the feasibility of catheter-based intravascular ultrasound (IVUS) to obtain quantitative and qualitative information of arterial anatomy in vivo and in vitro.i-7 Real-time 2-dimensionalultrasound imageshave detected plaque composition2 postangioplasty dissection,132and thrombus6 In vitro imaging of postangioplasty segmentshas demonstrated plaque morphology and qualitative characteristics of plaque disruption.l Human in vivo data have demonstrated an excellent correlation between IVUS and quantitative angiography for assessmentof peripheral arterial dimensionsin normal and minimally diseasedsegments.2The purpose of this study was to determine the feasibility and applicability of IVUS to assessthe acute effects of coronary interventional technologies.

C

METHODS

Patient group: Seventy coronary interventions were prospectively evaluated in 65 patients. There were 48 men and 17 women (mean age 63 f 6 years, range 49 to 76). The indication for intervention was progressive or unstable angina in all except 1 patient who presented with an acute myocardial infarction. Preprocedure angiographic stenosiswas 175% by visual estimate in all patients. Eleven patients had undergone previous balloon coronary angioplasty. Interventions performedwere: 45 balloon percutaneous transluminal coronary angioplasties (PTCAs), 11 directional coronary atherectomies,9 excimer lasers (4 of which were performed with adjunctive PTCA), 3 rotational atherectomiesand 2 stainlesssteel stent implantations. Coronary interventions were performed in a stanFrom the Department of Medicine, Division of Cardiology Duke Uni- dard manner over a O.OlCinch guidewire through a versity Medical Center, Durham, North Carolina. Manuscript received 9Fr giant lumen-guiding catheter or a 9.5Fr or 11Fr May 21, 1991;revisedmanuscriptreceivedand acceptedJuly 5,199l. Address for reprints: Charles J. Davidson,MD, Box 31195,Duke guide catheter for patients undergoing atherectomy. University Medical Center, Durham, North Carolina 27710. The size of the interventional devicewas selectedbased INTRACORONARY ULTRASOUND

1305

on visual estimateof the angiographicahynormal vessel diameter using the guide catheter for calibration. The sites of the 70 lesions were 29 right coronary arteries, 31 left anterior descendingcoronary arteries, 8 left circumflex coronary arteries and 2 saphenousvein grafts. All lesionswere within the proximal or midsection of the vessel. Ultrasound imaging: The coronary catheter used was a 4.8Fr monorail over-the-wire catheter (Boston Scientific Corporation, Watertown, Massachusetts). The catheter was advancedover a O.OlCinch guidewire immediately after the coronary intervention. A 20 MHz single piezoelectriccrystal transducer is mechanically rotated within the catheter sheath and attached to a diagnostic imaging console(Diasonics, Inc., Milpitus, California). Two-dimensional images were displayed on a video monitor of 512 X 5 12 pixel matrix that discriminated 64 shadesof gray. Images were recorded on 0.5-inch videotape for subsequent quantitative and qualitative analysis.Quantitative analyseswas performed off-line with resident software from a HewlettPackard SONOS 1000 system (Andover, Massachusetts). If a dissection was present, the area within the true lumen and behind the flap was planimetered.Passageof the catheter-basedultrasound was successfulin all patients in which it was attempted. There were no complications attributable to IVUS imaging. In patients 1 to 6, the software configuration of the ultrasound system was not calibrated for quantitative determination. Postproccdurecomputer-derivedquantitative measurementswere obtained in 23 consecutive arterial segmentsfor the image displaying minimal absolute lumen diameter and cross-sectionalarea. Due to the vast discordanceof the 2 techniques,no further angiographic quantitation was obtained. Cross-sectional area was obtained by planimetry. Excellent intraobserver and interobserver correlations have been previously reported.* Qualitative analyseswere evaluated by 2 observers unaware of the results of angiography. Vesselmorphology analyseswere undertaken for plaque composition (hard versus soft), plaque topography (eccentric versus concentric) and vesseldissection. Plaque compositionwas evaluated in a similar manner to that previously suggestedby Gussenhovenet a1.3 Hard plaque was defined as vesselwall thickening with bright reflected echoes or distal acoustic shadowing consistent with calcium, or both. Soft plaque lacked these characteristics and was representedas less dense echoesof thickened arterial wall. Eccentric plaque was defined as involving <50% of the total circumferenceof the vessel. Dissection was defined as separation of plaque from intima or media with encroachment of plaque onto the catheter blank or with independentmo1306

THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 68

tion from the vessel wall. Equivocal dissections were confirmed by hand injection of contrast media through the guide catheter during ultrasound imaging. If a dissection was present, contrast was visualized within the true lumen and in the dissectionplane. Contrast angiography: Contrast cineangiography in orthogonal views was obtained before and after the intervention with hand injection of nonionic contrast. Imageswere acquired in a 7-inch image intensifier mode at 30 frames/s. Quantitative digital cineangiographic images were analyzed with a previously validated, semiautomated edge-detection algorithm8 in the same patients measured with ultrasound. The guide catheter was used for calibration. Absolute lumen diameter and cross-sectional area were determined at the same position as the coronary angiogram image displaying minimal lesion diameter and area. Qualitative analysis of cineangiographic images were undertaken by consensusof 2 observersunaware of the angiography for determination of vesseldissection. Dissection was defined as either a linear filling defect within the vesselor an extrahuninal opacity.g Data analysis: The correlation between angiographic and intravascular ultrasound images was determined with linear regression. Categoric data were analyzed with a Breslow chi-square. RESULTS Quantitative data: The absolute lumen diameter and cross-sectionalarea for 23 consecutive postinterventional arterial segmentsdetermined by angiography and ultrasound were poorly correlated (r = 0.28, standard error of the estimate = 0.52 mm, p = 0.2, and r = 0.1, standard error of the estimate = 1.0 cm*, p = 0.7; respectively) (Figure 1). In general, IVUS measurementsof vesselsize were larger than with angiographic determinations. Qualitative data: Although variability in image quality existed, MIS provided morphologic information in all cases.Most lesionsconsistedof a mixture of both hard and soft plaque. Plaque composition was classified as hard in 30 (43%) lesions and soft in 64 (91%) lesions. Evaluation of plaque topography revealed that eccentric lesions, comprising <50% of the vessellumen, were observed in 34 lesions (49%), and concentric plaque was present in 36 lesions (51%). In the 70 interventions qualitatively evaluated by IVUS and angiography, vesseldissectionwas noted by ultrasound in 29 segments(41%), and in only 14 segments (20%) by angiography (xl,* = 9.8; p x0.05) There was concordance between the 2 techniques in 67% of sites.However, coronary dissectionswere noted

NOVEMBER 15, 1991

sound despite an adequate angiographic result. Thus, the smooth vesselcontour noted with silhouette contrast angiography appearsto underestimatethe extent of residual atheroma (Figure 4). In the evaluation of intracoronary stainless steel stent deployment, conventional angiography is limited in its ability to visualize relatively radiolucent struts. IVUS readily defined the strut and vesselarchitecture (Figure 5). Although this patient incurred a major dissection after conventional PTCA, catheter-basedultrasound demonstrated resolution of dissection without persistent flap or luminal compromise.

by ultrasound at 19 sites (27%) in which they were not detected with angiography. Figure 2 demonstratesan example of a right coronary artery PTCA in which hazinessbut no obvious dissectionis present at the dilation site. A large mobile intimal and medial dissection is present by ultrasound. Despite the ultrasound appearance, this patient had no acute adverse clinical events, and remains asymptomatic 7 months afterPTCA. Angiography identified 4 dissectionsthat were not detectedby ultrasound. Theseoccurred in vesselswhose postintervention lesion diameters by angiography were <2.5 mm. The luminal flap noted by angiography was probably “tacked up” by the near-occlusiveultrasound catheter. In the 8 patients undergoing excimer laser angioplasty, hard plaque comprising >50% of the lesion was noted in 4 patients and predominantly soft plaque was observedin 4 patients. An example of soft plaque ablation is shown in Figure 3. Vessel dissection was noted in 2 of 11 patients (18%) undergoing directional atherectomy and in 26 of 45 patients (58%) undergoing conventional balloon PTCA (p
DISCUSSION This study describesa series of patients in which IVUS was used to assessthe acute effects of interventional technologies. Intracoronary ultrasound appears feasible and safe for postinterventional imaging. The advantageof this technique over contrast angiography is in its ability to define postinterventional lesion characteristicsincluding lesion composition,topography and dissection. It has been previously shown in postmortem studies that plaque morphology including topography and lesion composition are determinants of pathologic outcome after coronary angioplasty.l” With IVUS, plaque

Minimal Diameter

0.12 0.10 -

.

m m..

xFIGURE 1. Intravascular

Minimal Area

.

E 0.08- .-: am . s *

ultrasound

raphy (Angto) far dete&ination of lumen diametsr (Diam) &ftJ and cross-sectional area (right). SEE = standard error of the estimate.

g 0.06 2 v) 3 0.04 2

r=0.28 SEE=0.52mm p=o.2 I

0

1

I

.

0.00

5

.

- m

0.02 -

I

2 3 4 Angio Diam (mm)

.

.

. I

I

I

r=O.i SEE=l.Ocm’ p=o.7 I /

0.00 0.02 0.04 0.06 0.08 0.10 0.12 Angio Area (cm 2,

FIGURE 2. Left, tight coronary arhwy angiogram immediatefy atter pwahiwus transiuminalcoromuyangiqhBy.Hahessifbnoted at dilation sites, but no obvious dissection farrow). Right, large intimal and medial dtssectlons noted wlthh area of hard plaque.

INTRACORONARY ULTRASOUND

1307

may assist in interventional device selection or prediction of long-term outcome. Thrombus could not be definitively identified in any patient. Previous studies evaluating thrombus and soft plaque have demonstratedthat their ultrasonic characteristics are often similar.3*6After peripheral venousangioplasty, IVUS has been demonstrated to be more sensitivethan angiography for detection of intraluminal thrombus as well as dissection.1t Coronary dissections were more often identified with ultrasound than with cineangiography. It has been well documented in previous studies that angiographic morphologic description, particularly dissection,is a significant predictor of ischemic outcome after PTCA.9J2 Despite a large dissection noted in the patient in FigFIGURE 3. lmmediiteiy after 2 passes with excimer lassr. ure 2, no adverseclinical sequelaewere noted acutely, Channels and sites of tissue ablation noted within soft plaque and the patient remains asymptomatic 7 months after (arrows). PTCA. Whether the improved sensitivity of ultrasound composition and topography were readily identified. to detect dissection impacts on therapeutic approaches Plaque composition was classified as hard or soft based to improve either acute or long-term outcome after anon previously suggestedpathologic validated criteria.3 gioplasty remains to be determined. As opposedto previous studies of noninterventional With the present state of IVUS image quality, more extensive characterization of lesion composition is not peripheral arterial segments,2discordance exists bevalidated and may represent an overinterpretation of tween IVUS and quantitative angiography for determithe images. However, when compared with angiogra- nation of postprocedurecoronary dimensions. Numerphy, the improved lesional descriptors with ultrasound ous factors may account for this discrepancy. A previ-

FtGURE 4. Leff, left anterior descending wrmary artey after directkmal coronary atherectomy. Nete smooth cantow of vessel at site of intervention (arrow). Right, corresponding intravascular uttrasound image demonstrating channel of excised plaqus (7 o’dodc) and resldwl athtwoma (12 o’dod&

FfGURE 5. Left, after major diswctbns

with

stent deployed into mid-k& ant&or art~.SbUtSZWC!!SMll ~WcoroMly peody with angiography (arrow). Right, stnrts readily visuallled wlthout evfdence of

1308

THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 68

NOVEMBER 15, 1991

ous study in dogs has indicated a poor correlation of ultrasound with angiography after PTCA.13 These investigators determined that an increase in measured vessel eccentricity accounted for the differences, and that ultrasound may be superior to cineangiography in defining vesselsize when lumen shapeis distorted. Tobis et all4 also demonstratedthat IVUS and quantitative angiography are discordant after balloon angioplasty. Preliminary data from our laboratory indicate that technical factors inherent to IVUS also contribute to the discordanceobservedwhen comparedwith angiography.15Eccentric catheter position, vesselcurvature and beam divergence are all potential sourcesof error in evaluating vessel size and echogenicity. Thus, both angiography and ultrasound may exhibit propertiesthat account for the discrepanciesobservedin quantitative comparisons. The size of the population presented in this study does not allow for definitive conclusionsregarding the differential acute effects of various interventional technologies.Rather, this study servesto describethe ability of IVUS to provide a more sophisticated and detailed evaluation of coronary interventions. The clinical use of this new imaging modality in the setting of noninvestigational interventions remains to be determined. Although quantitation of postintervention lumen dimensions may have limitations, the qualitative abilities of IVUS should improve the understanding of the acute results and therapeutic impact of new interventional devices when compared with conventional PTCA. REFERENCES 1. Tobis JM, Mallery JA, GessertJ, Griffith J, Mahon D, BessenM, Moriuchi M, McLeay L, McRae M, Henry WL. Intravascular ultrasoundcross-sectionalarterial imaging before and after balloon angioplasty in vitro. Circulation 1989;

80x873-882. 2. Davidson CJ, Sheikh KH, Harrison JK, Himmelstein SI, Leithe ME, Kiss10

KB, BashoreTM. Intravascular ultrasonographyversusdigital subtractionangiography: a human in uioo comparisonof vesselsize and morphology.J Am CON Cardiol 1990;16:633-636. 3. GussenhovenEJ, EssedCE, LanceeCJ, Mastik F, Van EgmondFC, Reiber J, BoschH, Van Urk H, Roelandt J, Born N. Arterial wall characteristicsdetermined by intravascular ultrasoundimaging:an in vitro study.J Am Co[l Cardiol 1989;14:947-952. 4. Yock PG, Linker DT, Angelsen AJ. Two-dimensional intravascular ultrasound: technical developmentand initial clinical experience.J Am Sot Echo 1989;2:296-304. 5. PandianNG, Weintraub A, Kreis A, Schwartz SL, Konstan MA, Salem DA. Intracardiac, intravascular,two-dimensional,high frequency ultrasoundimaging of pulmonary artery and its branches in humans and animals. Circulation 1990;81:2007-2012. 6. PandianNG, Kreis A, Brockway B. Detection of intra-arterial thrombus by intravascular high frequencytwo-dimensionalultrasoundimaging in vitro and in viva studies.Am J Cardiol 1990;65:1280-1283, 7. Harrison JK, Sheikh KH, DavidsonCJ, Kiss10KB, Leithe ME, Himmelstein SI, Kanter RJ, BashoreTM. Balloon angioplasty of coarctation of the aorta evaluated with intravascular ultrasound imaging. J Am CON Cardiol 199Q15: 906-909. 6. Mancini GB, Simon SB, M&hem MJ, LeFree MT, Friedman HZ, Vogel RA. Automatedquantitative coronary arteriography: morphologicandphysiologic validation in vitro of a rapid digital angiographic method. Circulation 1987;75:452-460. 9. Ellis SG, Roubin GS, King SB, DouglasJS, Weintraub WS, ThomasRG, Cox WR. Angiographic and clinical predictors of acute closure after native vessel coronary angioplasty. Circulation 1988;77:372-379. 10. Farb A, Virmani R, Atkinson JB, Kolodgie ID. Plaque morphology and pathologic changesin arteries from patientsdying after coronary balloon angioplasty. J Am Co11 Cardiol 1990;16:1421-1429. 11. Davidson CJ, Newman GE, Sheikh KH, Kiss10K, Stack RS, Schwab SJ. Mechanismsof angioplastyin hemodialysisfistula stenosesevaluatedby intravascular ultrasound. Kidney Int 1991;40:91-95. 12. Black AJR, Namay DL, Nicderman AL, Lembc NJ, Roubin GS, Douglas JS, King SB. Tear or dissectionafter coronary angioplasty: morphologiccorre lates of an ischemiccomplication. Circulation 1989;79:1035-1042. 13. NissenSE, Grines CL, Gurley JC, Sublett K, Haynie D, Diaz C, Booth DC, DeMaria AN. Application of a new phased-arrayultrasoundimagingcatheter in the assessmentof vascular dimensions.Circulation 1990;8 1:660-666. 14. Tobis JM, Mallery J, Mahon D, LehmannK, Zalesky P, Griffith J, GessertJ, Moriuchi M, McRae M, Dwyer ML, Green N, Henry WL. Intravascular ultrasoundimagingof humancoronary arteries in viva. Circulation 1991;83:913-926. 15. Sheikh KH, Trahey GE, Kiss10KB, Kiss10I, Davidson CJ. Fundamental limitations of intravascular ultrasoundin the assessmentof vascular size,gcometry, and echogenicity(abstr). J Am Co11 Cardiol 1991;17:233A.

INTRACORONARY ULTRASOUND

1309