352

tr v

r

Detection

t r

sis at the Site ofs

i n Anglographically Normal or Minimally NarrowedCoronary SegmentsMASAKAZU YAMAGISHI, MD, KUNIO MIYATAKE, MD, JUN TAMAI, MD,SATOSHI NAKATANI, MD, JUN KOYAMA, MD, STEVEN E. NISSEN, MD, FACC*

Osaka, Japan and Cleveland, Ohio

Objectives . The purpose of this study was to ose intravascularultrasound

g to examine the presence of occult athe

e-. at the site of focal v in ang aphic y normal or

minimally narrowed segments, t the role of atherosclerosisIn the development of v

.Background. Previous clip

and experimental studies havesu

ed that early ather k is is present at the site of fvrs

. However, no clinical data exist demonstrating occultdisr~~se 0 the site of v

m at

i

phically insignificant

M

, Twenty-two patients with chest pain at rest or duringexertion, or both, were studied . Vasospasm was provoked byIn administration of ergonovine m ate (0.01 to0.04 mg) . After relief of vasospasm by nitroglycerin administra-tion, intravascular ultrasound im

was performed with a 32.t, 20-MHz, synthetic aperture array ultrasound

device.Rest. F

v

(arterial diameter reduction a-90%)with ST-T segment elevation was provoked in 15 patients: in the

Coronary vasospasm plays a pivotal role in provoking myo-cardial ischemia in patients with variant angina, myocardialinfarction and sudden death (1,2) . This abnormal coronaryvasoconstriction can occur at angiographically normal orminimally narrowed sites as well as at significant stenoticsites (3-6) . In cases where vasospasm is superimposed on anangiographically evident lesion, the atherosclerosis is con-sidered to be present at the site of vasospasm (3,4) . How-ever, the existence of atherosclerosis at the site of vaso-

From the Cardiology Division of Medicine, National CardiovascularCenter, Osaka, Japan and 'Cleveland Clinic Foundation, Cleveland, Ohio .This work was supported in part by grants from the Japan Heart Foundation,Tokyo (Dr. Yamagishi) and the Takeda Medical Research Foundation, Osaka(Dr. Y ishi) and a research grant (3A-7) from the Ministry of Health andWelfare of Japan, Tokyo (Dr. Miyatake). This work was presented in part atthe 42nd Annual Scientific Session of the American College of Cardiology,Anaheim, California, March 1993 .

Manuscript received April 26, 1993 ; revised manuscript received August9, 1993, accepted September 15, 1993.

Address for correspondence: Dr. Masakazu Yamagishi, the CardiologyDivision of Medicine, National Cardiovascular Center, 5-7-1 Fujishiro-dai,Suita, Osaka 565, Japan .

01994 by the American College of Cardiology

JACC Vol. 23, No . 2February 1994 :352-7

CORONARYARTERYDISEASE

left anterior descending coronary artery in 8 patients and in theright coronary artery in 7 . The remaining seven patients (controlgroup) showed diffuse narrowing, averaging 22 t 12% (mean tSD) in diameter from the baseline angi rams after ergonovineadministration. Atherosclerosis, defined as a significantly thick .ened inti leading edge (0.42 t 0 .07 mm) associated withincreased sonolucent zone (0.57 t 0. ), was detected byultrasound at all 15 sites with f vases m, although these siteswere normal or minimally narrowed by a iog phy. In contrast,seven segments t the control group exhibited a thin intimalleading edge (0.14 t 04 mm, p < 0 .01) and sonolucent zone(0.10 t 0.07 mm, p < 0.01), indicating the absence of localizedathe

erotic lesions .Conclusions. These results indicate that atherosclerosis is

present at the site of f v spasm, even in the absence ofangiographically significant coronary disease . We suggest that theexistence of such athe lerotic lesions is related to the occur-rence of focal v s m in the clinical settings.

(J Am Coll Cordial 1994;230352-7)

spasm in angiographically normal segments remains unclear(5,6) because angiography can image only the silhouette ofthe vessel lumen and may miss intramural disease .

A necropsy study in patients with sudden death hasshown focal intimal thickening at the coronary site wherevasospasm had been documented before death (7), suggest-ing that the presence of occult atherosclerosis was closelyrelated to the occurrence of vasospasm. This hypothesis wassupported by Shimokawa et al. (8), who showed that occultatherosclerosis was invariably found at the site of histamine-induced focal vasospasm in miniature swine . However, inthe clinical setting, it has remained difficult to demonstratethese occult lesions in angiographically normal or minimallynarrowed segments where vasospasm was documented .

Recent advances in the intravascular ultrasound imagingtechnique have enabled visualization of the intraluminal andintramural morphology of the coronary arteries (9-12) .However, few clinical data exist with regard to the ultra-sound findings at the site of the vasospasm (13) . Thus, in thepresent study, we studied patients with or without focalcoronary vasospasm in normal or minimally narrowed cor-

0735-10971941$7.00

JACC Vol. 23, No . 2February 1994 :352-7

Figure 1 . Representative coronary an-giographic and ii.travascular ultra-sound images and schematic represen-tation in a patient with focal vasospasmin the right coronary artery (Patient 1,Table 1). A, Before ergonovine admin-istration, there were minor lumen ir-regularities at the mid-right coronaryartery (arrow) . B, At 3 min after intra-coronary ergonovine (0.01 mg), thissegment was severely narrowed to 90%stenosis (arrow) . C, Intracoronary ad-ministration of nitroglycerin (0 .25 mg)relieved the stenosis, and this arterybecame angiographically normal (ar-row) . D and E, At the site of vasospasm(arrow in B and C), the intimal leadingedge and sonolucent zone were typi-cally thickened, indicating the pres-ence of a circumferential atheroscle-rotic lesion .

onary arteries using intravascular ultrasound imaging todetermine the presence or absence of atherosclerosis at thesite of vasospasm.

MethodsStudy subjects . From February 1992 to December 1992, a

total of 22 of 528 patients who had diagnostic coronaryangiography were entered into the present protocol . All 22patients had chest pain without reproducible positive find-ings on exercise testing . Although not a consecutive series,the study group represented all recruitable patients who gavewritten informed consent . The chest pain occurred at rest in16 patients and both at rest and during exertion in 6 patients .Although 14 patients showed ST-T segment elevation>0.1 mV on the electrocardiogram (ECG) during spontane-ous angina, ECG changes were not detected in 8 patients .There were 20 men and 2 women aged 40 to 69 years (meanage 57 years) .

Provocation and imaging protocol . This protocol wasapproved by the Ethical Review Committee of the NationalCardiovascular Center (No . 4-3, 1992) . All antianginal drugsexcept sublingual nitroglycerin were discontinued at least12 h before catheterization . The catheterization procedureswere performed in the postabsorptive state with administra-tion of 3,000 U of heparin . Coronary angiography wasperformed by the standard Judkins technique (517, Softip,Schneider) using a biplane cineangiography system (Bi-Coroskop, Siemens) .

The absence of lumen narrowing >25% in diameterreduction was confirmed by the initial injection of contrastmedium (iopamidol, Schering Japan) into the left main andright coronary arteries . Subsequently, I ml of ergonovinemaleate solution (0 .01 mg) was injected into either the leftmain or right coronary artery through the catheter . We

YAMAGISHI ET AL .INTRAVASCULAR ULTRASOUND IN CORONARY SPASM 353

selected the arteries for the injection of ergonovine byreferring to the ST-T segment changes during spontaneousattacks. In patients who did not show any ST-T segmentchanges, ergonovine was first injected into the left maincoronary artery and then into the right coronary artery .Ergonovine (0 .01 mg) was administered every 3 min untilvasospasm was provoked or the maximal dose of 0 .04 mghad been given (14) . A standard 12-lead ECG was carefullymonitored and recorded every 30 s .

Focal vasospasm was defined as a localized transientvessel narrowing associated with myocardial ischemia, in-cluding ST-T segment cllevat~jn >0 .1 mV by the ECG . Whenchest pain associated with ST-T segment elevation wasexperienced, angiography was immediately performed and,subsequently, nitroglycerin (0 .25 mg) was injected into thecoronary artery through the catheter to relieve spasm (Fig .1) .

The ultrasound system used in the present study has beendescribed in detail elsewhere (12,15) . Briefly, the intravas-cular ultrasound imaging device consisted of either a 32- or64-element transducer mounted at the tip of a catheter witha diameter of 3.5F or 5 .017 (Cath-Scanner, Endosonics) .Ultrasound images were synthesized and displayed using a512 x 512-pixel matrix with an 8-bit gray scale (256 levels) .The overall echo dynamic range of the system was 80 dB .In combination with a catheter transducer, the dynamicrange was 40 to 55 dB, depending on the patient and tissueinterfaces . Ultrasound gain settings were adjusted for op-timal visualization of the lumen-intima boundary (19 to26 dB), and lumen images were acquired without frameaveraging.

A 7F guiding catheter (Superfiow, Schneider) was usedfor insertion of the intravascular ultrasound catheter . Afteradditional administration of heparin (5,000 U), a 0 .014-in .(0.036 cm), very flexible, steerable angioplasty guide wire

354

YAMAGISHI ET AL .INTRAVASCULAR ULTRASOUND IN CORONARY SPASM

(USCI) was advanced through the ultrasound catheter andfluoroscopically guided into the distal coronary artery . Theultrasound imaging catheter was then advanced over theguide wire into the vessel . The dimensional measurementaccuracy, imaging characteristics and safety of this catheterdesign have been previously reported (15) .

In patients with focal vasospasm, an imaging probe wasinitially placed within a site proximal to the vasospasm andwas then carefully advanced to observe the site wherevasospasm had occurred. Injection of a small amount ofcontrast medium often assisted the recognition of the intimalleading edge border of the site examined . This injection ofcontrast medium also provided a method to ensure thatinsertion of the ultrasound catheter probe had not reinducedcoronary vasospasm . After imaging a site distal to thevasospasm, the probe was carefully pulled back through thevasospasm site, again confirming the presence or absence ofany lesions .

In patients without focal vasospasm, the ultrasound cath-eter was advanced into the coronary artery as distally aspossible and was then carefully pulled back to continuouslyimage the wall morphology. All images were recorded onS-VHS videotape for subsequent data analysis .

U1 measurements . The frames of the cardiaccycle that showed the largest size of the lumen with optimaldelineation of the blood-intima border were selected forsubsequent measurement . Measurements were performed atthe sites where the vessel wall was maximally thickened(Fig. 1). The maximal thickness of the intimal leading edgewas taken from the brightest echo signal innermost to thetransducer, whereas the adventitia was identified as the nextbright layer extending about the circumference of the vessel .

The maximal thickness of the intimal leading edge wasmeasured by placing an electronic cursor at the borderbetween the vessel lumen and the intimal leading edge and atthe trailing edge bordering the sonolucent zone . We alsomeasured the maximal thickness of the sonolucent zone asthe distance from this trailing edge to the leading edge of theadventitia (Fig. 1). However, a trilaminar structure is some-times absent in normal subjects, as reported by others (16) .In such cases, we considered the site to be normal if therewas no evidence of luminal encroachment. Measured valuesrepresent the average of 3 consecutive beats .

According to previously reported findings (15), we de-fined a lesion as atherosclerotic if the thickness of the intimawas >0.3 mm or that of the sonolucent zone >0 .2 mm, orboth. We defined fibrocalcific lesions as the presence of highintensity echoes with acoustic shadowing . There was goodinterobserver (r = 0.97, y = 0.98x - 0.04, SEE = 0.04 mm)and intraobserver (r = 0 .98, y = 1 .05x - 0.01, SEE =0.03 mm) reproducibility for measurement of the intimalleading edge for 10 randomly selected sites . Angiographicresults were evaluated by at least two independent observersnot associated with the present study .

Statistics. All data are expressed as mean value ± SD .We used the unpaired Student t test to compare groups of

patients with and without focal vasospasm . We used thepaired t test when the data were compared within eachgroup. We considered differences significant if p < 0 .05 .

Results

Provocation of f, k • v palm. In the total group of 22patients examined, vasospasm (90% to 100176 in diameterreduction) associated with ST-T segment elevation wasprovoked in 15 patients (Patients 1 to 15, Table 1) :n the rightcoronary artery in 7 patients (Fig . 1) and the left anteriordescending artery in 8 (Fig . 2). Fourteen of 15 patientsshowed ST-T segment elevation in the same lead that wasrecorded during the spontaneous anginal attacks, and Ipatient exhibited significant ST-T segment elevation onlyduring the ergonovine provocative test . The mean age of thisgroup was 56 ± 7 years . After nitroglycerin Administration,the sites of vasospasm returned to an angiographicallynormal state in eight patients and minimally narrowed inseven .

Ergonovine did not provoke focal vasospasm in sevenpatients (Patients 16 to 22, Table 1) who averaged 60 ± 9years in age (p = NS compared with patients with spasm) . Inthese control patients, the coronary arteries were diffuselynarrowed (22 ± 12% diameter reduction) after ergonovinewithout ST-T segment deviation and contained no lumenabnormalities after nitroglycerin (Fig . 3).

Intrav ular ultrasound images . In all patients with focalvasospasm, there was significant thickening of the intimalleading edge associated with an increased sonolucent zone atthe site of vasospasm (Fig . I and 2) . At these segments, themean value of the maximal thickness of the intimal leadingedge was 0 .42 ± 0.07 mm and that of the sonoluc^,nt zonewas 0.57 ± 0 .30 mm (Table 1) . At sites proximal or distal tothe segment with vasospasm, ultrasound imaging showed athin intimal leading edge and sonolucent zone in adjacentsegments in 13 of 15 patients . In two patients, however,there was a slightly thickened intimal leading edge andsonolucent zone just proximal to the spasm site . A fibrocal-cific lesion, characterized by the presence of acoustic shad-owing in the ultrasound image, was observed at only I of the15 vasospasm sites .

In contrast, of the seven patients without focal vaso-spasm, none exhibited focal thickening of the intima . A thinintimal leading edge and sonolucent zone were observed,although these ultrasound layers sometimes could not beobserved circumferentially within a single frame (Fig . 3) .The mean value of the intimal leading edge was 0.14 ±0.04 mm and that of the sonolucent zone was 0 .10 ± 0.07 mm(Table 1). These values were compatible with those innormal subjects examined by the same ultrasound deviceand previously reported in detail elsewhere (15) . There wereno complications during the catheterization procedures .

JACC Vol . 23, No . 2February 1994:352-7

JACC Vol. 23, No . 2February 1994-352-7

Table 1 . Summary of Angiographic and Intravascular Ultrasound Studies in 22 NtOrts

DiscussionPrevious _,Audies. The precise pathophysiology of coro-

nary vasospasm remains uncertain, particularly with regardto the extent and severity of underlying atherosclerosis . Inpatients with mild fixed lesions on angiography, spasm oftenoccurs at the site of minimal obstruction, suggesting arelation between early atherosclerosis and clinical vaso-spasm (6) . However, spontaneous or ergonovine-provokedspasm sometimes occurs at sites with no evidence of athero-sclerosis on angiography (1-6) . In patients with suddendeath or myocardial infarction who had coronary vasospasmdemonstrated by angiography, necropsy findings typicallyindicate focal thickening of the intima at the site of vaso-spasm (7,17) . Accordingly, the presence of occult athero-sclerosis at angiographically normal sites could be a signifi-cant factor in the occurrence of focal vasospasm in theclinical setting. However, except for a preliminary study(13), few data exist demonstrating occult lesions at the sitesof vasospasm .

Advantage of the present study. Irtravascular ultrasoundimaging is particularly valuable in detecting the presence ofatherosclerosis in angiographically normal or minimally nar-rowed coronary segments (10-12,15). Therefore, in thepresent study, the site of focal vasospasm was examined byintravascular ultrasound and compared with findings in a

YAMAGISHI ET AL .

355INTRAVASCULAR ULTRASOUND IN CORONARY SPASMal

*p < O.W . tp < 0.01, EM = ergonovine rpaleate; F = female ; I = irregular ; IL = intimal leading edge ; LAD = left anterior descending coronary artery ;M = male ; N = normal ; NTG = nitroglycerin ; RCA = right coronary artery ; SN = sonolucent zone ; ST T = ST segment elevation .

control group. We found a thickened intimal leaditub edgeand an increased sonolucent zone by ultrasound at everyspasm site, even if the segment was angiographically normalor minimally narrowed . Previous validation studies using themultielement ultrasound system have demonstrated thatatherosclerosis is represented by the presence of either athickened intimal leading edge or an increased sonolucentzone (12,15). Accordingly, the present findings demonstratethe universal presence of atherosclerosis at the site of focalvasospasm .

With few exceptions, sites proximal or distal to thesegments with vasospasm where the excessive vasocon-striction did not occur after ergonovine, contained no signif-icant lesions by ultrasound . The findings that these adjacentsegments do not respond to ergonovine may establish thatlocal hyperreactivity at the site of vasospasm is produced byatherosclerosis. Kaski et al . (18) have demonstrated byangiography that the site of the vasospasm has a hyperreac-tive response to ergonovine compared with the response ofother "normal" segments . A study in experimental animalshas also shown that occult atherosclerosis can produce localhyperreactivity of the coronary artery (19) . Therefore, theoccult atherosclerosis observed in the present study is themost likely source of local hyperreactivity and ergonovine-induced spasm .

PtNo.

Age(yr)/

Gender EM (mg)StudiedVessel

ST-T SegmentChanges After EM

- Angiography Ultrasound (mm)

% Diameter ReductionAfter EM

DiameterAfter NTG IL SN

I 531M 0 .01 RCA ST T (11, 111, aVF) 90 N 0.43 0 .632 471F 0 .01 LAD ST T (V^) 99 1 0.44 1 .523 60/M 0 .01 RCA ST T (11, 111, aVF) 100 N 0.46 0.724 651M 0 .01 RCA ST T (11, 111, aVF) 99 1 0.38 0.525 671M 0 .02 RCA ST 1 (11, 111, aVF) 100 N 0.52 0.426 62/.M 0 .01 LAD ST T (V 3-V 5 ) 90 0 .32 0.747 551M 0 .04 LAD STT NO 99 0 .42 0.348 60/M 0 .01 LAD ST T (V,-V 5 ) 100 0 .53 0.449 401M 0 .03 LAD ST t (V,-V,) 90 N 0 .34 0.2810 49/114 0 .03 LAD ST t (V 3-V6 ) 90 N 0 .31 0 .3811 60/M 0 .01 RCA ST T 01, 111, aVF) 100 N 0 .42 0 .5612 53/M 0 .01 RCA ST T (11, Ill, aVF) 90 1 0 .38 0 .4413 54/M 0 .02 LAD ST T (V 2-V,) 100 1 0 .46 0 .4814 62/M 0 .03 RCA ST T (11, ill, aVF) 99 N 0 .36 0 .4215 52/M 0 .01 LAD S1, T (V,-V,) 90 1 0 .46 0 .64

Mean t SD 56 ± 7 0 .02 ± 0.01 0.42 ± 0 .07 0 .57 ± 0 .30

16 56/M 0 .04 RCA No change 39 N 0 .18 0 .1117 69/1VI 0.04 RCA No clLiagc 18 N 0 .16 0 .1018 67/M 0.04 LAD No change 38 N 0.10 019 59/F 0.04 LAD No change 12 N 0.12 0 .1920 561M 0.04 LAD No change 8 N 0.16 0 .0921 43/M 0.04 LAD No change 26 N 0.10 0 .0422 69/M 0.04 RCA No change 16 N 0.18 0 .12

Mean ± SD 60±9 0.04* 22 ± 12 0 .14 t 0 .04t 0 .10 ± 0 .07t

356

YAMAGISHI ET AL.INTRAVASCULAR ULTRASOUND IN CORONARY SPASM

It is intriguing to consider the underlying pathologyrepresented by a thickened intimal leading edge or increasedsonolucent zone, or both, in the ultrasound images of thevessel wall . However, it remains unclear whether ultrasoundobservations are closely related to actual pathologicchanges, such as lipid-laden plaque or medial thickening .Ultrasound images of the vessel wall reflect acoustic prop-erties but not the precise histology of the vessel wall (20) .Quantitative tissue characterization (21) may be helpful tocorrelate ultrasound images with the actual histopathologyof the vessel wall .

. The present data may provide use-ful insights into the mechanism of coronary vasospasm in theclinical setting . First, focal vasospasm occurred at the site ofatherosclerosis regardless of the angiographic findings .Thus, it is likely that angiographically normal segments withfocal vasospasm contain occult atherosclerosis that cannotbe demonstrated by conventional angiography (3) .

Advanced atherosclerotic lesions accompanied by calci-fication (9-12) were demonstrated in only one patient, whereasthe remaining patients had relatively early-stage atherosclero-sis by ultrasound. In the experimental model, microscopicexamination of the spasm site has shown lesions consistingmainly of cell proliferation without fibrocalcific changes, sug-

sting that the occurrence of vasospasm is closely related tothe presence of early atherosclerosis (19). The present dataco that early atherosclerosis plays a pivotal role in theoccurrence of vasospasm in humans as well as animals . Thereremains the possibility that spasm itself might contribute toaccelerated progression of atherosclerosis (22) .

JACC Vol . 23, No. 2February 1994:352-7

2. Angio phic and intravas-cular ultrasound findings in the patientwith focal vasospasm in the proximalleft anterior descending coronary ar-tery (Patient 2, Table 1) . A, Beforeergonovine administration, there wereminor lumen irregularities at the prox-imal let, anterior descending coronaryartery (arrow). B, At 2 .5 min afterin ronary ergonovine (0 .01 mg),this segment was severely narrowed to

stenosis (arrow). C, Intracoronarynitroglycerin administration (0 .25 mg)relieved the narrowing .. and there re-mained lumen irregularities at the siteof vasos m (arrow) . D, Intravascularultrasound image at the site proximalto vasospasm (arrow D in C) exhibiteda thin intimal leading edge and a thinsonolucent zone (arrow) . E, At the siteof vasospasm (arrow E in C), there wasa thickened intimal leading edge (ar-row) and a thickened sonolucent zone(h: : arrows) that represented a non-circumferential atherosclerotic lesionfrom the I to 8 o'clock positions . F,The site distal to vasospasm (arrow Fin C) did not exhibit signs of athero-sclerosis.

Figure 3. Angi phic and intravascular ultrasound findings in thesubject without focal vasospasm in the right coronary artery (Patient16, Table 1). A, Before ergonovine administration, there were nolumen irregularities in the right coronary artery . B, After intracor-onary ergonovine (0.04 mg), this artery was diffusely narrowed to a390A reduction in diameter. C, Intracoronary administration ofnitroglycerin (0.25 ) relieved the diffuse n wing. and thisartery became angi phically normal (arrow). D, Intravascularultrasound examination did not reveal any abnormal thickening ofvessel wall along this artery (arrow) .

JACC Vol. 23, No . 2February 1994 :352-7

Study limitations. Several limitations of the present studyshould be noted. We did not compare the present findingswith those in patients with classical effort angina with whomfocal vasospasm was not provoked . Therefore, it is stillunclear whether the occurrence of vasospasm depends on anintrinsic property of atherosclerosis . Because we used evgonovine maleate to provoke vasospasm, one might arguethat the provoked spasm may not coincide with spontaneousspasm. However, 14 of our 15 patients with focal vasospasmshowed ST-T segment elevation in the same leads duringspontaneous and ergonovine-induced attacks . Moreover, aprevious report (23) has demonstrated that ergonovine-induced vasospasm is often similar to spontaneous vaso-spasm.

Although the current study demonstrated occult athero-sclerosis in a relatively small number of patients with vaso-spasm, the findings esutblik the role of early atherosclerosisin the occurrence of vasospasm in clinical patients . Furtherintravascular ultrasound studies will be necessary to exam-ine the functional disintegration of the diseased segment incombination with pharmacologic testing (24) .

We thank Susan D . Sekowski for excellent secretarial assistance .

ReferencesI . Hillis LD, Braunwald E . Coronary-artery spasm . N Engl J Med 1978 ;299 :

695-702.2 . Maseri A, Severi S, De Nes M, et al . Variant angina : one aspect of a

continuous spectrum of vasospastic myocardial ischemia. Am J Cardiol1978-,42:1019-35 .

3 . Heupler FA . Syndrome of symptomatic coronary arterial spasm withnearly normal coronary arteriograms . Am J Cardiol 1980,45 :873--si .

4 . MacAlpin RN . Relation of coronary arterial spasm !o sites of organicstenosis . Am J Cardiol 1980,46:143-53 .

5 . Mercurio P, Kronzon 1, Winer H . Spasm of a normal or minimallynarrowed coronary artery in the presence of severe fixed stenoses of theremaining vessels: clinical and angiographic observations . Circulation1982-,65 :825-30 .

6. Bertrand ME, Lablanche JM, Tilmant PY, et al . Frequency of provokedcoronary arterial spasm in 1089 consecutive patients undergoing coronaryarteriography. Circulation 1982 ;65:1299-306 .

7. Roberts WC, Curry RC, Isner JM, et al . Sudden death in Prinzmetalsangina with coronary spasm documented by angiography . Analysis ofthree necropsy patients . Am J Cardiol 1982 ;50 :203-10.

YAMAGISHI ET AL .

357!NTRAVASCULAR ULTRASOUND IN CORONARY SPASM

8. Shimokawa H, Tomaike H, Nabeyama S, et al . Coronary artery spasminduced in miniature swine : angiographic evidence and relation to coro-nary atherosclerosis . Am Heart J 1985 ;110 :300-10 .

9. Pandian NG, Kreis A, Brockway B, el al . Ultrasound angioscopy:real-time, two-dimensional, intraluminal ultrasound imaging of bloodvessels. Am J Cardiol 1988 ;62 :493-4 .

10 . Yock PG, Johnson EL, Linker DT . Intravascular ultrasound . Develop-ment and clinical potential . Am J Card Imaging 1988 ;2:185-93 .

11 . Mallery JA, Tobis JM, Gfiffith J, et al. Assessment of normal andatherosclerotic arterial wall thickness with an intravascular ultrasoundimaging catheter. Am Heart J 1990 ;1191392-

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

13 . Kijima M, Komatsu N, Owada T, Ikeda K, Maruyama Y . Intravascularultrasound in patients with vasospastic angina pectoris [abstract] . I AmColl Cardiol 1992 ;19 Suppl A :286A .

14 . Hackett D, Larkin S . Chierchia S, Davies G, Kaski JC, Maseri A .Induction of coronary artery spasm by a direct local action of ergonovine .Circulation 1987 ;75 :577-82 .

15 . Nissen SE, Gurley JC, Grines CL, et al . Intravascular ultrasoundassessment of lumen size and wall morphology in normal subjects andpatients with coronary artery disease . Circulation 1991 ;84 :1087-99 .

16. Fitzgerald PJ, St Goar FG, Connolly AJ, et al . Intravascular ultrasoundimaging of coronary arteries . Is three layers the norm? Circulation1992 ;86 :154-8 .

17 . EI-Maraghi NRH, Sealey BJ . Recurrent myocardial infarction in a youngman due to coronary arterial spasm demonstrated at autopsy . Circulation19901:199-20-1,

18. Kaski JC, Maseri A, Vejar M, Crea F, Hackett D, Halson P . Spontaneouscoronary artery spasm in variant angina is caused by a local hyperreac-tivity to a generalized constrictor stimulus . J Am Coll Cardiol 1989 ;14 :1456-63 .

19. Egashira K, Tomoike H, Yamamoto Y, Yamada A, Hayashi Y, Naka-mura M. Histamine-induced coronary spasm in regions of intimal thick-ening in miniature pigs : roles of serum cholesterol and spontaneous orinduced intimal thickening . Circulation 1986;74 :826-37.

20. Kawano S, Yamagishi M, Nakatani S, et al. Comparison of intravascularultrasound images of human coronary arteries obtained by mechanicalrotating and synthetic aperture array systems [abstract] . J Am CollCardiol 1993 ;21 Suppl A:449A .

21 . Sudhir K, Fitzgerald PJ, Yock PG . Potential for tissue characterization ofplaque and arterial wall using intravascular ultrasound . SPIE 1993 ;1894 :115-21 .

22. Nobuyoshi M, Tanaka M, Nosaka H, et al. Progression of coronaryatherosclerosis : is coronary spasm related to progression? J Am ConCardiol 1991 ;18 :904-10 .

23 . Curry RC, Pepine CJ, Sabom MB, Conti R . Similarities of ergonovine-induced and spontaneous attacks of variant angina . Circulation 1979;59 :307-12.

24 . Yamagishi M, Nissen SE, Booth DC, Gurley JC, Fischer C, DeMaria AN .Impaired nitroglycerin induced vasodilation in coronary atherosclerosis :evidence from intravascular ultrasound [abstract] . J Am Coll Cardiol1992 ;19 Suppl A :95A .

I'll