Which angiographic variable best describes functional status 6 months after successful single-vessel coronary balloon angioplasty?

Which angiographic variable best describes functional status 6 months after successful single-vessel coronary balloon angioplasty?

Angiographic Vatia After Successful Single-Vessel V&i& BENNO J. RENSING, PIM J. DE FEYTER, Ratrerdom. MD, MD, WALTER FACC, R. M. HERMANS. PATR...

665KB Sizes 10 Downloads 79 Views

Angiographic Vatia After Successful Single-Vessel V&i&

BENNO

J. RENSING,

PIM J. DE FEYTER, Ratrerdom.

MD, MD,

WALTER

FACC,

R. M. HERMANS.

PATRICK

W. SERRUYS,

MD, JAAP MD, PH,,,

VY. iX FACC

The Netherlmds

Ubjechws. The Fiat of (his study was to determine which qunntltative angiographic variable hes!describesfunctional s&tas 4 mnnths alter coronary balkon angiaplasty. B&@mmd. Several angiographic resten& criteria have been developed. Thesecan be divided into (hose that describe the chge in lesion severity and lhose that merely describe lesion severity at follow-up mgiogtnphy. The functional signikance of thesecriteria ls un!umwn. MeUn&. We studied 350 paients with single-ves%l cmmwy artery dim who underwent a singk-site bakmn dilation. Sm. tivity and specilicity curw were const~ckd I& thp @i&on of rqhml status md en&se ekctrowdkgrapby of four quanUtalive att&rqhic variahl~ that describe resknosis. The point of high& diqnmlic amtracy for the variaMes was dekrmined al the i&rsectktl of the semitivily nodspxikity curves. Rf!adts of

exerciseekamurdiagnphywereeansidered indkattve for ischemia6 monthsafterar@@sty if t~rizentalor dowoskping ST segment deprasim el mmaaxrrcd.

for dkgnostk aeeursq varisbles were oblrdned ers of ischemia suggsls or inereose iit hion severity in majw epkardkl VW& at ahi& wonary flow reserve is unabk to lneet myowdkI dcrrulsds.

Soon after the introduction of percutaneous ttansluminal coronary balloon angioplasty the vexing ptobtem of restenasis became apparent. Approximately 30% to 40% of patients will show functional or an&graphic signs, or both, of restenosis in the tist months after a successful angioplasty procedure. A multitude of angiographic reslcnosis criterk have been developed over the last I2 years (1). Restenosia criteria currently in use ELUI be divided In those that describe ihe change in lesion severity Lam the postangiop&y situation up to the follow-up angiognm and those tha? meraly describe lesion severity at follow-upangiography. An example of the first category is the loss in lumen diameter of X1.72 mm as proposed by Serrttys et al. (Z), and ai -Y-‘~ of the latter category is the criterion of >50% diameter stenosis at follow-up. Criteria that describe a change in lumen diameter may ignore the functional significance of

the lesion at follow-up, especially in large vessels (3), whereas criteria that only describe the situation at follow-up

Manuscript received December 17, 1991; revised manuscript received July 6, 15% accepted July 14. 15% Addrcrsrorcoricrwndence: Patrick W. S:m~ys, MD, PhD. Cntheterizs lian L&raton,Thordxccnter. Erasmur University. P.O. Box 1738.3M DR Rottcrdam. ‘be Nethtrlandr.

severity (change in minimal lumen diameter, change in percent diameter stenosis) and variables that merely describe lesion severity at follow-up angiography (miniinn lumen diameter and percent diameter stenosis).

to

, *n.q,.”

will preselect lesions with a suboptimal result after angioplasty and thereby disregard the magnitude of the reactive

intimal hyperplasia [I). Recurrence of a flow-limiting stenesis can usuallybe identified by symptoms of chest pain similar to those that occurred before angioplasty. In addition to the medical history, exercise electrocardiogtaphic (KG) testing is generally performed as a noninvasive approach to con&m the recurrence of a coronary artery obstruction t,ecause it is a relatively simple, safe and inexpensive test. To determine which quantitative angiogaphic variable best predicts the functional status of the individual patient, we studied the recnrrenee of anginal complaints and positive ECG exercise test results 6 months after successful anginplasty in a selected patient group with single-vessel disease and single-site dilation. The functional variables (recurrence of angina and a positive exercise test result) were correlated with quantitative angiogmphic variables of change in lesion

57.07. (67 of IBll 28.7Tc (25 of $7) 38.7% (14 of 62)

of the gain achieved at angioplasty WHLBI 4); 5) increase of lesion severicy tc‘ 2-B?% diameter stenosis at follow-up: and 6) deterioration in minimal lumen diameter of ~0.72 mm from immediately postangioplasty to follow-up. The latter criterion is based on the long-term variability of minimal lumen diameter measurements using the CAAS system (0.36 mm). This variability is of the mean difference of two measurements of the same lesion filmed at two catheterization sessions an average of 90 days apart 15). This long-term variability reflects the long-term random variation in lesion measurements from coronarv aneioenms made at different catheterization sessions us& the 2AA.S sy~tcm (5). The use of SD would include 68.3% of the variability. while the use of 2 SD (2 x 0.36 = 0.72 mm) includes 95.5% of the variability.

0.91’

I $3

I

Results Table I summarizes the baseline characteristics of the 350 study patients. No differences were found in the proportion of patients with recurrent angina and ST depression at exercise with respect to the vessel dilated (Table 2). The occurence of Q waves in the area supplied by the dilated artery as an indicator of pnor transmural infarction (Table 2) was low. Recurrent an@na and qumtitative an’&raphy. At follow-up 102 (29%) of 350 patients had recurrent angina. Percent correct classitka!ion Fif recurrence of angina (sensitivity) and percent correct classikation of absence ofangina at fallow-up (specificity) as a function of cutoff points for the different quantitative angiographic variables are given in Figure I. The point of intersection of the sensitivity and specificity curves represents the cutoff point for which diagnostic accuracy was best. The static variables of minimal lumen diametct follow-up and percent diameter stenosis at follow-up performed equally well with a sel4ivity and specificity slightly >70% at cutoff points of I .45 mm and46.5%, respectively. Variables that reflect the change in lesion severity from immediately postangioplasty to follow-up performed on!y slightly less favorably. vritk a sensitivity and specificity of just <7&% at cutoff points of -0.30 mm for the change in minimal Idmen diameter and

at

-10% for the change in percent diameter stenosis. To compare the diagnostic accuracy of the different variables, receiver operator characteristic (ROC) curves were EMIstrttcted (Fig. 2). Quantitative angiographic variables denoting a change in lesion severity arc only slightly less accurate than those that denote a static measurement of lesion severity at follow-up. Pasitive exereisa test and quantitative angiopphy. The exercise lest result was abnormal in I16 patients (35%). Percent correct classification for an abnormal test Wnsitivity) and percent correct classification for a normal test (specificity) as a function of cmoff points for the different quantitative angiographic variables are given in Table 3. It is clear that the diagnostic accuracy of exercise testing was lower than for anginal status at follow-up. The optimal combination of sensitivitv and saecificitv was 44% fo: all angiographic variables, with the’static vsiables performing slightly better the variables of change (Table 3). How&r, the cuto~Tpoints associated with the point of interscction of ihe sensitivity and specificity curYes were similar to those obtained with anginal status. This was 45.5% diameter stenosis at follow-up, 1.46 mm minimal lumen diameter at follow-up, achange of 10% in diameter stenosisand a change of 0.32 mm in minimal lumen diameter. The RDC curves for the different quantitative angiographic variables are shown in Figure 3. fkurrenr angina* pnsitiwe exercise test MUR and mterk ask criteria in current me. Tables 4 and 5 list the sensitivity. specificity and predictive values for recurrent angina and a positive exercisetes( result of different previous:yproposed

than

diameter stemsis (%)

restenosis criteria. None of the criteria predicted recurrent angina and positive exercise test results with great accuracy. In particubs a posirive exercise lest result was very hard to predict with all angiographic restenosis criteria. Criteria that require a large change INHLBI 1 and the 20.72mm triterion) and the NHLBI 2 criterion that requires an extraordinary high diameter stenosis at follow-up had a low sensitivity. In fact, the NHLBI 2 crikrion is more a predictor of total occlusions because 22 of 26 lesions that fulfilled this criterion were totally occluded at follow-up. Among the 330 patients who underwent exercise testing, 19 of the 22 lesions that fulfilled the NHLBI 2 criterion were totally occludd at follow-up. The NHLBI 3 and 4 criteria and the 250% diameter stenosis criterion Performed better. The 50% diameter stenosis cutoff point lies close to the optimal cutoffpoint of 46.5% and is therefore one of the best predictors of recurrent angina or a positive exercise test r&It whereas the NHLBI criterion (change in diameter stenosis ~30%) and the ~0.72~mm criterion are clearly remote from the optimal cutoff points of - IW and -0.32 mm, respectively (Fig. 1, C and D, Table 3). Diagnostic aecuraey of quantitative anglography tn large Venus small vessels. Coronary artecies taper from proximal IO distal. The comparison of proximal with distal coronary

1

Figure I. Percent correct cla5rilication of recurmx of angina (sensitivity [&us] %, darker curve) and percent correct classification of ab sence of angina at follow-up (specificity [Spec] %, lighter curve) as a function of cutoff points for the different quantitative angiograpbic vtiables. The point of intersection of the two wwes denotes the cutoff point with the hiiest diaauodc accuracy. Where appropriate. the 50% diameter stenosis and the 20.72.mm loss in minimal lumen diameter [MLD] r&enosis criteria were drawn in the figure. A, Curves for minimal lumen diameter at follow-up. 8, Curves for percent diameter stenosis at f&w-up. C, Curves for change in minimal lumen diameter at follow-up. D. Curves for change in percent diameter stenosis (DS) al follow-up.

Figure 2. Receivaoperalorchamcleristic furverforcomparison of the diagnostic accuracy of an&al stagesat follow-up for minimal

Minimal lumen diameler bum) Diamelerstenosis (%) Change in minimal lumen &am&3 (InIn, Change in diameter

stenc5is(%) *Paintat which sen&ivity equalsspecificity.At lhepoint of intersection. dimrustic accuracy is rlaximal and sensitivity= speclficily. vessels carries the disadvantage of grouping together vessels of diierent diameter. To study whether the quantitative angiogragbiL values found apply io bo::h large nnd small vessels we determined the points of intersection of the sensitivity and spci$city curves for all four measuremems of restenosis. The vessels were therefore divided into two equally large groups according to fhe reference diameter so that each g~oup contained 50% of the study ressels (Table 6). Large -JesseIs had a reference diameter 22.63 mm (n = 176) and small vessels had a reference diameter ~2.63 mm. The point of intersection of the sensitivity and specificity curves of the absclute diameter at follow-up was the only quantitative angiographic cutoff point that was diierent in large vessels as compared with smaller vessels.

Diiussion Patienl sekctian and methodologic considerations. In this

study we preferentially studied patients with single-vessel disease and with a single lesion that was successfully dilated. In these patients only restenoais of this lesion can be held responsible for an abnormal ECG response at follow-up exercise testing or recurrent angina. whereas in multivessel disease the responsible lesion is not always easily idcntifiable. Moreover, coronary angioplasty in multivessel disease will not result in complete revascularization in a considerable number of cases (IO). With conventional exercise protocols, ECG leads and ECG criteria, exercise testing is characterized by a high specificity and a moderate sensitivity (Ii). Furthermore, its sensitivity increases with the extent of coronary artery disease, which implies a low sensitivity in patients with single-vessel disease (12-14). This explains the low predictive accuracy found in our population of patients with singlevessel disease for the diierent restenosis criteria. These findings are comparable with the findings of Bengtson et al. (IS), who found a sensitivity of 32% and 9 specificity of 79% for a positive exercise ECG. It is known that diagnostic accuracy for restenosis of recurrent angina is better than for ST segment change (15,16). From the data of Benglson et al. (15). a sensitivity for recurrent angina of 5% and a specificity of 73% can be

Figure 3. Receiver

opaator characteristic wrwfor comfison

of

the diagnostic xcuracy of exercise electrocardiogmphy for minimat lumen diameter at followup (kavy S&I Ike)+ percent diameter stznkr at f&w-up (lk& z&d IW. change in minimaI lumen diameter (dasknl Ibe with sqa& and change in percent diameter stenosis (d.Wed liw with Merkks).

calculated. They applied the 50% diameter crMon for restenosis. Zaidi et al. (17) reported a sensitivity of 70% and a specificity of 66% for recurrence angina as a test for restenosis. Although the predictive accuracy of quantitative angiographic variables was generally poor in the present study. it is remarkable that the points of intersection of sensitivity and specificity curves were similar for two different markers of myocardial ischemia. Angioplasty of the left anterior descending artery made up 55.4% of all procedures in this study. It might be argued that the IaEe mass of mylcardium supplied by this artery, which is potentially iscbemic in case of severe renarrowing. would render the tindings of this study applicable only to left anterior descending artery lesions. However, no dlerences were found in the proportion of patients with recurrent angina and ST depression at exercise whh respect to the vessel dilated (Table 2). prior myocardial infarction is known to falsely increase the accuracy of exercise testing (18). However, the occurrence of Q waves in the area supplied by the dilated artery (Table 2) was low, indicating only a small possible influence on our findings. Angiographic r&en&s and funclid st&us. Reslenosis after a successful angioplasty procedure is now viewed as a fibroproliferative repair process in response to traumatic injury to the vessel wall (19,20). We recently showed that lumen narrowing after angioplasty Occurs to a certain exten! in all dilated lesions (21) and that angiographie resrenosis is the tail end of a normally distributed phenomenon. The restenosis rate is then dependent on the cu10iT criterion applied. Generally two types of angiographic restenosis criteria have been developed: criteria that denote the change

the

322

RPNSlNO ET AL. ANO,mK*PHY AND

T&k

4. Diagnostic

Accuracy

lSCHEMlA

AFTER

ANGlDPLASTY

of Different Angiegraphic RestenosisCriteria for Patients With Recurrent Angina CrilUiOll

Spccificily PVP WN

95.6% I237 of 248) 12.5% (29 of 40) 76.5% (237 oi 310)

NHLB, 2

NifLBi 3’

NHLBI 4.

250% DS

20.72 mm

7.4% (26 of 350) 20.6% (21 of 102) 98% I243 of 248.1 80.8%

31.2% (MY of 30) t4.8W (62of 102) 81% 12w of 247) 56.9% (62of IWi 83.3% (2@3OE240)

34.7% (121of 149) &.7% (66of 1112) 77.7% [I92 oil411 54.5%

31.7% (Illof 63.7% (65 of 102, 81.5% (202of 248) 58.6%

17.7% (62 Of3501 lY.2% &loOf lo?) 91.1% (226 of 248)

166of 121) 84.2% (I92 of 2181

(65 of 84.5% (202or 239)

(21 of 26) 77.3% (243 of 314)

in stenosis severity at follow-up angiography (e.g., the r0.72-mm change criterion) and criteria that assess stenosis severity at follow-up angiography (e.g., the 150% diameter stenosis at follow-up criterion). From a functional aoint of view restenosis can be detected by recurrence of an, +a and by several noninvasive tests. These tes!: ?re anaed at detecting myocardial ischemia due to a flow-limiting

stenosis

artery and give no informaticb on the magniiude of the lumen-narrowing process in the individual lesion. An&graphic restenosis criteria that give a static assessment of lesion severity at iolow-up have the disadvantage of presetecting lesions with a marginal angioplasty result (I). This means that these lesions have to undergo only a small deterioration to cross the cutoff point and be classified as “restenosed.” Classically the 50% diameter stenosis criterion is applied to classify lesions or patients as “restenosed” at follow-up angiography after angioplasty. This definition is historically based on the physiologic concept of coronary gow reserve introduced by Gould and others (22) in 1974 and is taken because it represents the approximate value in animals with normal coronary arteries at which in an epicardial

TaMe 5. Diagnostic Accuracy of Dierent at Follow-Up

1III

61.5% of 62) 78.5% (226ol 2%)

I40

blunting of the hyperemic response occurs. Although the 50% diameter stenosis criterion is attractive because it links the angiographic appearance of a lesion with the clinical situation of the patient, it tells us nothing about the dynamic behavior of the restenosis process. Our findings underscore the significance of the 5056 diameter stenosis criterion because the optimal cutoff point for prediction of functional status 6 months after coronary angioplasty was found to be close to this value (46.5% diameter stenosis). However, diagnostic accuracy of ihe absolute stenosis diameter at 6 month follow-up was similar to percent diameter stenosis (Fig. 2 and 3) with a point of intersection of the sensitivity and specificity curves for both angina1 status and exercise test results at approximately I.45 mm. This observation indicates that ao absolute measure of stenosis severity is equally predictive of clinical status after angioptasty as a relative measurement. These values correspond well with the findings of Wilson et al. (23). They found that coronary Bow reserve dropped below 3.5 (the lower threshold of normal) at a minimal cross-sectional area of I.5 mm2 and a percent area stenosis of 75%, which corresponds to a mini-

Angiograptlic Restenaais Criteria for Patients With a Positive Bicycle Ergometry Response ClifUtOil

Characteristic Prevalence of reslenoris Scnsilivity Specificily PVP PVN

NHLBI

1

11.2% (37 of 330) 14.7% 11701116) 9C.15 ‘!94 d 214) 45.% (17 of 371 66.2% (1Y.iof 293)

NHLBl2 6.7% (22 of 130) 9.5% (II of 1161 94.% (203 of 2141 sD9F (II or221 BS.Y’% no3 Of3081

NHLBI 3’ 30.7% (101 of 329) 37.9% (44 of 116) 73.2% (I56of213) 43.6% (44 Of 101) 68.4% (156of 2281

NHLBI 4’

a5G%DS

zo.72 mnl

34.7% I4 of 3291 46.5% (5: of 116) 71 .a% (IS3 al 2131 41.4%

31.2% I103 of 3343) 41.4% 148 of I lb) 74.3% (I59 of 214) 46.6%

17% (S6 of 3301 26.1% (3lafll6) 88.3% (I89 or 214) 55.4%

(54 of 114) 71.2% I153 or 215)

I48 of 1031 7o% (159of 227)

(31 Of56) 6% ,189 of T14)

(I


Change in Mismd

-0.41 (II = 169;

-0.17 mm tn = 174

-0.30 In = 1611

mat lumen diameter of 1.4 mm and a percent diamctcr stenosis of 50%. respectively. Wijns et al. (24) demonsimted a steep increase in pressure drop over Left anterior descending artery stenoses once a Critical value of minimal crosssectional atea of 2.5 mm* was reached. This corresponds to d minimal lumen diameter of 1.78 mm. The pressure measurements were made with a dilation catheter across the stenosis (cross-sectional areaaf catheter = 0.64 mm’). If this value is subtracted from the 2.5mm* value, a minimal Lumen diameter value of 1.54 mm emerges, which is close to the 1.45 mm found in the present study. An approach that more closely retlects the magnitude of the reactive intimal hyperplasia after angioplasty is applying criteria that describe the change in lesion severity at follow-up angjography. The major critique of this type of criterion is that it might disregard the functional significance of a lesion at follow-up. For instance a lesion with a postangioplasty percent diameter stenosis in the range of0 to 15% can undergo a large deterioration and still not he flow limiting, However, our study shows not only that the static variables of minimal lumen diameter and percent diameter stenosis perform equally well in predicting .clinical significance of a lesion 6 months after successful caronary angiap!asty, but also that the variables of change in lumen diameter and change in percent diameter stenosis were only slightly less accurate in predicting the clinical significance of the lesions (Fig. 2 and 3). Therefore, variables of change. apart from their usefulness in reflecting the magnitude of the reactive hvmrolasia. also reflect nearlv to the same extent as the static vatiables,‘the clinical signif;cance of the lesion at follow-up. The optimal cutoff point for the variables of changewas -0.36mm for the changein lumen diameter and -10% for change in percent diameter stenosis with nearly equal diagnostic accuracies showing that absolute change in lesion severity (in mm) and relative changein lesiou severity (in percent) perform equally well in the prediction of recur-

.. .

Lumen mameter

-0.z I”nl I” 7 ,761

rent angina or a positive exercise ECG response 6 months after coronary angioplasty. Liiitatious uf the &rQ_ Fin:, arc.-e sophisticated invasive and noninvasive methods are available for the assessment of the functional significance of a coronary stenosis. It is known that exercise thallium scintigraphy has a higher diagnostic accuracy than does ECG exercise testing (25). The present data originate from a multicenter trial and therefore it is logistically diicuh to srandardiie the methodology of radionuclide exercise tests or flow reserve measurements in all panicipatirtg centers. Exercise ECG testing on the other hand is inexpensive and safe, and identical exercise protocols can be easily implemented in the participating centers. As proposed by Popma et at. (26). paired stress tests (shortly after angioplastyand at 6 monthsfollow up) should ideally be obtained, otherwise the interpretation of au ischemic exercise test at foltow-up may be more ditfrcult, especially in patients with multivessel coronary artery disease. Our study group consisted of patients with single-vessel disease in whom complete revasculatization was achieved. However, the absence of myocardial ischemia at hospital discharge after angioplasty was not objectively confirmed by exercise testing. The diagnostic accuracy of quantitative angiographic variables for the prediction of recurrent angina and an abnormal ECC response at exercise was not very high. Nevertheless, the absolute values of sensitivity and specificity were not crucial to this study, hut rather tbe point of intersection of the sensitivity aad specificity curves. Angina! medication was not standardized in this study. This factor might have influenced the assessment of functional status; however, an attempt was made to withdraw antianginal medication at least 24 h before exercise testing. Finally, all angiograms were preceded by an intracoronary dose of nitrates and not all patients were using vasodilator drugs at the time of exet-cise testing, This factor might have shifted the points ofintersection toward a higher

minimal

lumen

diameter

and a lower

percent

diameter

ste-

nosis.

number of patients studied and the fact that the same optimal values for diagnostic accuracy of the various quantitative am&graphic variables were obtained for the prediction of two diierent markers of ischemia (angina1 status and ST depression at exercise) suggest that these values reflect the lesion severity or increase in lesion severity in major epicardial vessels at which coronary flow reserve is unable to meet myocardial demands. Relative measurements (percent diameter stenosis) and absolute measurements (minimal lunen diameter) were found to be equally predictive of isI .lemia. Because the minimal lumen diameter is the most txxnbiguous measurement of lesion severity (independent of an arbitrary normal part of the artery), this measmc can be a more reliable sumgate for clinical outcome than the classic percent diameter stenosis measurement in the many restenosis prevention trials with drugs and new devices currently underway or in the design phase. Conclusions.

The

large

Tbc technical asririance oflaap Pamc~er is aratefullv acknwkdwd.

References I. Beatt KJ, Sermys PW, Hugeoholtz PG. Reslenorir after conwary snsio_ p&sly: new standards for clinical studies. I Am Coil Cwdiol 199o:Is: 491-a. 2. Serruys PW, Luijtcn HE, Beau Kl, el al. fncidencc of restcnosis after successful coronary angioplaty: a time related plmomenon. A quantidative angiwaphic study in 342 consecutive patients at I.&3 and 4 months. Circulation 19@71:361-71. 3. C&t? RM, Fortin DF, Ftid Dl. et al. Resteno$i$ atIer comaary anJo. plasly: an overview. I Am Coil Cardial 1991:17:2l%13B. 4. SCMYE J’W. Rutsch W, Heyndrickx GR, et al. Fwention afrestcnorir

--__, 6. Reikr JHC, Sermys PW, Slager CJ. Quantilative coronary and IelI vcnticular cioean&gmphy. Melhodology and clinical application. In: Reiber IHC. Strays PW, cds. State of the Art in Qunmlimtive Coronary Angiogmphy. Dordrecht: Martinus MhoiT, 19%x162-89. 7. Rsiber JHC, Sermys PW. Quantitative angiopphy. In: Marcus ML, Schelbwt HR. Skorton DJ. WolfGL, eds. Cardiac Imaging. a Companion to Braunwalds Hca~ Disease. Philadolohia: WB Saunders. 1991:211-Ro. 8. RriberJHC, Sermyr FV, KooyntanCi Slager N, SchuurtiersJHC,den Boer A. Approaches toward standardization in acquisition and quantitathn n?artrri*! dimm+w frnm cin&n_~r. In Ref 6:145-55 9. Sermys PW. Deckers JW, Luiten HE, et al. Lao8 acting coronary

rasdiktmy actionof the molrldamine met&ok Sin I: a quanlitadve angiographic study. EM Hem I 1987$.:26~70. !h Semys PW, Foley DP. deFe3 ter PI. Reatenosir alter coronary an& pksty: a prcfmsal of new campamtive approaches based an quantitative angiognpby. &Heart J 1992$3&l-24. IO. Samson M, Meester HI. de Fefler PI. Straws BH. Sermys PW. Succcssf~~lmultiple wgmcnt cc~ronaryan&&sty: efkctofcampkkncss of revascularization in single-vessel multilesions and multivtsuls. Am Heart J 19!%120:1-12. R, Salcedo E. Lmlbmnan 1. Day K. verse unassisted analysis of tk cxercisc dsctrwardiogram in patients without myaardial infarction. 1 Am Coil Cardial 1%77:10:794-9. 12. Martin CM. McConahay DR. Maximal lreadmill exercise electrwurdia 8ra~by: camlatianr wilh cormm,y arteriography and cardiac bcmodynamiu. C~rsulation 1972$6z956-62. 13. GofdschlagerN. Seltzer A. Cohn K. Trcadmillatress tests as indkatars of presence and icverity of coronary artery diaase. Ann Intern Mcd 1976$5231-86. 14. Ilijneke RD. Aswop CA, Talmw JL. Clinical significance of upsloping ST segmeals in excrckc electrocardiography. Ciiculation 1%30:61:671-8. 15. Bag&on JR, Mark DB. Honan MB, et al. Detection ofrertenorir after ckctive percutsnews iransluminal conmary angioplasty using the exe*cisc treadmill test. Am J Cardiol 19X$55:3-34. 16. Rosing DR. Van Radrn MJ, Mincemoyu RM. ct al. Exercise, ekctnr cardiagrapbk and func6onal responses after pcsutanews tnnrluminal cornnary ?.n&plasly. Am J Cardiol I9%;S3(suppl C):XC-4lC. 17. zaidiAR, HollmanJ.Galan K, et.& Predktive value dcbat direomfon for rertenosis lollwring successlul coronary angioplasty (abstr). Ciiuklion 1985;72:1 D-456. 18. Dctraw R, Simpfendolfer C, Day K. Cmnparison Ot strw digital verdricukSraphy, sins5 thaJJium scintigraphy, acd dighal Ruwwopy in the dipgnolis of conmary xtery diiw in subjects witbout prior myo. cardial infarction. Am J Cardiol 1985;56:434-40. 19. Ewd CE, Van den Brand M, Bscka AE. Tranrluminal conmary angiopksty and early restawis! fibroccllular oceluion after wall laceralion. Br Heart J 113:49:393-6. 20. Nobuyarbi M, Rimura I’, Ohishi H. ct al. Restcaasis after percutaoew~ translumioduaonaryangioplasty: p&&gicobrcwation5in28 patients. 1 A”t Co!, CardhI 1491:17:43L9. 21. Rensir@ BJ. Hermans WRM. Dtcken JW, ds Fcyter A, T@seo JGF. Sc~ys PW. Luminal narrowing after prculaneous transhtmiral cornnary an8iwlasty follows P near Gaussian dislnftion. A qumitalive atuiographk stiy in 144s successfullydilatedkion. I AtnCoU Catdiol lw2;19?939-4s. 22. Gould KL, Lipszomb K. Hamilton GW. PbySil~& basis for assessing critical stemw8: instantarYous Raw reyansc and re@onal dktniution during ccawry hypenmis as mcasurcs ofcoronary flow reperyt. Am I Cardiol lYM$3:87-94. 21. Wilson RF, Marcus ML. White CW. Prediction ot the physiolo8jc rignillcance of coronary ntterial lesions by quantitative ksicm 8eonwryin pstieols with limited coronary artery dixue. Ciiktion 1987:75:72&32. 24. Wdns W. Sc~yr PW. RcibcrlHC.etaJ. Qwmtitalivc ar&raphy ofthe IrR ant&n dcnreeding coronary attery: corr&ions with faewre gradient nnd resuhs of exercise thallium srinti8rapby. Circulation 1985; 71:273-9. 25 CalY RM, Ohnun EM, Fiid DJ. el rd. Restcmosis: the clinical buo. In: Topal BJ, ed. Texlbwk of lntewctttiattal Cardklc~y. New Vnk: WB Saunders, 1990:362-w 26. Popma JJ. &Iii RM. Tap01 El. Clinical trials of reaenosk following coronary an8ioplrty. Circulation lbl;M:142&36.

If.Dstrano

Computer assisted