impact of restenosis after optimal directional coronary atherectomy on regional left ventricular...

Q1997 by Excerpta Medica, Inc. 0002-9149/97/$17.00 545All rights reserved. PII S0002-9149(96)00813-2

/ 2w1a 0798 Mp 545 Monday Jan 27 04:55 PM EL–AJC (v. 79, no. 5 ’97) 0798

Impact of Restenosis After OptimalDirectional Coronary Atherectomy on

Regional Left Ventricular FunctionPierre Coste, MD, Serge Sempe, MD, Pierre Dos Santos, MD, Catherine Jaıs, MD,

Abdallah Moussari, MD, Francois Duclos, MD, Simone Bonoron Adele, PhD,and Pierre Besse, MD, FESC

To assess the effect of optimal directional coronary ath-erectomy (DCA) on restenosis and left ventricular (LV)function, 95 patients who underwent DCA and adjunc-tive balloon angioplasty for de novo lesions were pro-spectively followed for 6 months. Absolute and relativecoronary lumen measurements were analyzed with on-line quantitative coronary angiography. LV volumes,ejection fraction, and segmental wall motion were mea-sured off-line according to the radial method for LV ci-neangiograms acquired in a right anterior oblique pro-jection. Target vessels were the left anterior descendingartery in 63 patients and right coronary artery in 32.Mean ({ SD) reference diameter was 3.58 { 0.65 mm.Mean lumen diameter improved significantly after DCAfrom 1.19 { 0.44 to 3.03 { 0.45 mm, yielding a 14 {10% residual stenosis. Overall angiographic restenosis

rate (ú50% stenosis in diameter) at control was 23%. Inpatients without restenosis, there were no significantchanges in LV volumes or in LV pressures. In this sub-group, ejection fraction improved significantly in the leftanterior descending group (mean difference 3 { 10%,põ0.04). Moreover, there was an increase in fractionalshortening of all anterior segments (mean difference 11{ 16%, p õ0.005). Improvement in fractional shorten-ing was less marked in the right coronary artery groupeven without restenosis. We conclude that: (1) optimalDCA can achieve a low restenosis rate in selected largevessels, (2) long-term beneficial effects on regional LVfunction are possible, particularly in patients with leftanterior descending disease and in the absence of cor-onary restenosis. Q1997 by Excerpta Medica, Inc.

(Am J Cardiol 1997;79:545–552)

Directional coronary atherectomy (DCA) was firstproposed in order to reduce the incidence of re-

stenosis after balloon angioplasty1–3 but controlledtrials have failed to confirm a clear superiority ofatherectomy.4,5 However, a renewal of interest forDCA has been generated by the Optimal Atherec-tomy Restenosis Study which has shown lower ratesof restenosis (angiographic 29.6% and clinical18%).6 Consequently, such techniques should bringabout midterm improvement in coronary flow andtherefore restore ventricular function within themyocardial segments chronically exposed to isch-emia.7–12 This study prospectively evaluated the util-ity of optimal DCA in 95 consecutive patients whosystematically underwent 6-month follow-up angi-ography after a successful procedure. To investigatethe hypothesis that reducing the restenosis ratewould improve regional left ventricular (LV) func-tion, we compared the coronary angiogram course ofthe treated narrowing with that of cardiac functionestimated from segmental wall motion.

METHODSPatients: Between August 1992 and June 1995,

122 DCA procedures were performed in our center.Four were unsuccessful or gave way to complica-

From the Catheterization Laboratory and Intensive Care Unit, IFRCoeur-Vaisseaux-Thrombose, University of Bordeaux II, Bordeaux-Pes-sac, France. Manuscript received May 31, 1996; revised manuscriptreceived and accepted October 7, 1996.

Address for reprints: Pierre Coste, MD, Hopital Cardiologique,IFR Coeur-Vaisseaux-Thrombose, Avenue Magellan, 33604 Bor-deaux-Pessac, France.

tions (3.3%). Three patients (2.5%) developed myo-cardial infarction defined as new Q waves in the elec-trocardiogram and/or creatine kinase myocardialisoform more than twofold normal values. All thesepatients completed clinical follow-up, and 95 under-went angiography at 6 months, giving a 81% angio-graphic control rate in successful procedures. Thir-teen of the controlled patients (12%) initially had aslight elevation in creatine kinase above the upperlimits of normal. The remaining patients were eithersymptom-free or had Canadian Cardiac Society class1 or 2 angina with medical treatment. None requiredcardiac revascularization either by coronary bypasssurgery or repeat coronary angioplasty. Eligible pa-tients who all had symptomatic ischemic heart dis-ease were selected if the reference diameter of theselected vessel wasú3.0 mm. Otherwise, indicationfor DCA was determined according to the anatomiccondition of the targeted vessel and morphologic cri-teria of the lesion to be treated.1–5 Patients who pre-sented with restenosis after previous angioplastywere not included in this study.

Atherectomy procedure: DCA was performed withlarge catheters (size 7Fr in 87% of cases) accordingto the usual procedure to eliminate as much residualstenosis as possible.6,9 Enlargement of the vascularlumen was obtained through a stepwise approachthat comprised multiple cuts at low (10 to 20 psi)and then at high pressure (40 to 60 psi). Additionalangioplasty using high-pressure inflation with a non-compliant balloon was almost systematically used tocomplete the atherectomy. Mean ratio of balloon sizeover reference diameter never exceeded 1.1. A pre-

546 THE AMERICAN JOURNAL OF CARDIOLOGYT VOL. 79 MARCH 1, 1997


FIGURE 1. Evaluation of regionalwall motion according to the ra-dial method percent shorteningare expressed as mean { 1 SD.EF Å ejection fraction; S Å seg-ment.

dilation of the stenosis was necessary for 1 patientonly. All patients were given aspirin (250 mg/day)and intravenous heparin (10,000 IU bolus dose) be-fore DCA. Patients selected for follow-up had to ful-fill criteria for technical success defined by: (1) re-sidual stenosis õ50% in diameter in at least 2different radiologic projections; (2) absence of pro-cedural complications (vessel occlusion, myocardialinfarction, emergency coronary bypass surgery, re-peat angioplasty) during the hospitalization stay; and(3) no elevation in creatine kinase myocardial iso-form more than twofold normal values.

Quantitative analysis of angiograms: Coronary an-giograms were quantified using ACA software in-cluded in Philips Digital Cardiac Imaging system(Eindhoven, The Netherlands). The reproducibilityof measurements for this system has already beenreported.13 Two independent observers (P.C. andS.S.) performed these quantitative analyses withoutknowledge of LV function data. Coronary lesionswere assessed on end-diastolic images by analyzingthe same segment length after systematic intracoro-nary injection of vasodilator (Sin1; 0.5 mg). Condi-tions of centering, magnification, and radiologic pro-jections were carefully repeated during follow-upstudy. The guiding catheter was used as a calibrationdevice to express measurements in absolute values.Restenosis at the site of the preceding coronary ath-erectomy was defined by a renarrowing of the vessellumen ú50% in diameter on the control angiogram.

Hemodynamic data and quantitative evaluation ofventricular function: Baseline values of intraventricu-lar systolic and end-diastolic pressures, and aorticpressures were recorded with an external fluid-filledmanometer by averaging values from 10 cardiac cy-cles with a computerized system that processed thepressure curve signal.12 LV angiograms were filmedat 50 images/s, 15 minutes before atherectomy. TheLV cavity was opacified in the right oblique anteriorprojection (307) with 0.5 ml/kg of contrast mediumusing an automatic injector. Ventricular contourswithout ventricular premature beats only were takeninto account to evaluate segmental wall motion. LVcontours on end-diastolic and end-systolic frames

were manually traced by an independent observer.Contours were digitized and calibrated according toa computer program that takes into account distor-tion due to the image intensifier. LV end-diastolicvolume index, LV end-systolic volume index, andLV ejection fraction were calculated according toSimpson’s rule.12 Regional wall motion was ana-lyzed using a radial method developed by Ingels etal.12,14,15 Schematically, an index of wall displace-ment was measured for 10 angular sectors using acenter of contraction located at 69% of the axis dis-tance that joins the upper edge of the base of theaorta to the LV apex at end-systole. Then, segmentalwall motion was expressed as a percentage of theend-diastolic value of the radius of each sector (seeFigure 1). Normal values of ejection fraction andfractional shortening were calculated in a controlgroup of 25 patients without cardiac disease or sig-nificant coronary lesions.14 Normal range values(mean{ SD) were 70{ 6% for LV ejection fraction.The segments of the anterior wall, segments 6 to 10,showed the following percent shortening: S6, 35 {5%; S7, 52 { 6%; S8, 57 { 7%; S9, 59 { 7%; andS10, 52 { 5%. For inferior and basal zones, percentshortening of segments 2 to 5 were: S2, 20 { 7%;S3, 34 { 9%; S4, 38 { 8%; and S5, 31 { 5%. Toinsure better reproducible measurement conditions,patients were kept under the same pharmacologicregimen between the 2 examinations, and catheteri-zation was repeated according to the same protocol.Whenever feasible, patients had a programmed con-trol at 6 months (average 6 { 1.9). A few of themunderwent coronary angiography earlier than ex-pected because of recurrence of unstable angina, sug-gesting a rapid restenosis process.

Statistical analysis: Quantitative data are expressedas mean { 1 SD. Comparisons between groups weretested globally using analysis of variance for re-peated measurements. When significant differenceswere found, comparisons within groups for contin-uous variables were made using Student’s paired ttest. Between-group comparisons were undertakenwith Student’s unpaired t test for unmatched series.Categorical variables were compared using chi-

CORONARY ARTERY DISEASE/LV FUNCTION AND OPTIMAL DIRECTIONAL ATHERECTOMY 547


TABLE I Clinical and Angiographic Data of Study Population

Variables Right Coronary ArteryGroup (n Å 32)

Left Anterior Descending ArteryGroup (n Å 63)

Total Population(n Å 95)

Clinical dataAge (yr) 54 { 12 56 { 12 55 { 12Male gender 26 (81%) 53 (84%) 79 (83%)Hypercholesterolemia 25 (78%) 45 (71%) 70 (74%)Systemic hypertension 11 (34%) 31 (49%) 42 (44%)History of smoking 23 (72%) 45 (71%) 68 (72%)Diabetes mellitus 4 (12%) 13 (21%) 17 (18%)Family history of CAD 14 (44%) 31 (49%) 45 (47%)Prior myocardial infarction 16 (50%) 30 (48%) 46 (48%)Delay from myocardial infarction to DCA (d) 36 { 56 29 { 32 31 { 41Unstable angina 22 (69%) 50 (79%) 71 (75%)1-vessel disease 16 (50%) 39 (62%) 55 (58%)Multivessel disease 16 (50%) 24 (38%) 40 (42%)7Fr atherectomy device 28 (87%) 52 (83%) 80 (84%)

Quantitative angiographyMinimum lumen diameter (mm)Before procedure 1.29 { 0.50 1.14 { 0.40 1.19 { 0.44After procedure 3.14 { 0.50* 2.97 { 0.42* 3.03 { 0.45*Control 2.21 { 0.86‡ 2.15 { 0.72‡ 2.17 { 0.61‡

Reference diameter (mm)Before procedure 3.67 { 0.69 3.53 { 0.63 3.58 { 0.65After procedure 3.69 { 0.70 3.49 { 0.53 3.56 { 0.60Control 3.52 { 0.66† 3.44 { 0.58 3.47 { 0.61†

% Diameter stenosisBefore procedure 64.2 { 11.3 67.3 { 8.45 66.2 { 9.6After procedure 14.0 { 12.4* 14.8 { 9.5* 14.5 { 10.5*Control 37.1 { 20‡ 37.4 { 18.6‡ 37.3 { 18.9‡

Restenosis ú50% 9 (28%) 13 (21%) 22 (23%)

*p õ0.001 before versus after procedure; †p õ0.02; ‡p õ0.001 control versus after procedure.Numeric values are expressed as mean { SD, categorical values are numbers and percentage of the total number in each group.CAD Å coronary artery disease; DCA Å directional coronary atherectomy.

square analysis, or Fisher’s exact test when appro-priate. Statistical analyses were performed with SASsoftware (SAS Institute, version 6.04, Cary, NorthCarolina). A p level of probability õ0.05 was re-tained as significant.

RESULTSPopulation characteristics: The population was di-

vided into 2 subgroups according to the myocardialarea to be revascularized, and therefore according tothe targeted coronary vessel (i.e., right coronary orleft anterior descending artery). Baseline clinicalcharacteristics and angiographic data of the overallpopulation and subgroup analysis are detailed in Ta-ble I. These characteristics were similar to thosefound in studies dealing with DCA1–6 and, in addi-tion, they were similar in the 2 coronary territorysubgroups. Most patients presented with unstable an-gina, and 46 (48%) had experienced a recent myo-cardial infarction in the vascular territory of the tar-get vessel. This recruitment feature may be explainedby an anatomic selection based on the morphologiccharacteristics of the lesions, foremost of which wasintraluminal thrombus. Among the 46 patients withprior myocardial infarction, 41 had myocardial in-farction õ2 months before the procedure. Amongthese patients, 56% were initially treated with athrombolytic agent and coronary recanalization wasangiographically documentedõ6 hours after the on-set of chest pain (average 3.1 { 2.3 hours). The dis-

tribution of patients treated in the days after myo-cardial infarction was not different with respect tothe target vessel. All patients conserved contractilitywithin the infarcted area and all had normal coronaryflow (Thrombolysis In Myocardial Infarction trialgrade 3) in the infarct-related vessel. The DCA pro-cedure was performed after a period of clinical sta-bilization, ú5 days after an ischemic episode at restand ¢8 days after myocardial infarction.

Angiographic restenosis at six months: Figure 2 il-lustrates the immediate postprocedure and 6-monthoutcome of our population in terms of the degree ofstenosis measured by quantitative angiography.Twenty-two patients (23%) had angiographic reste-nosis. During follow-up, none of them experienceda new myocardial infarction. Ten of these 22 patientswere rehospitalized for recurrence of angina pectorisand underwent a new intervention for myocardial re-vascularization, either repeat balloon angioplasty (nÅ 7) or coronary bypass surgery (n Å 3). Therefore,only 11% of the entire population underwent a targetvessel revascularization in the territory concernedwith atherectomy. No other clinical cardiac eventwas noted during the follow-up period.

If one applies the commonly accepted criterion ofú50% stenosis in diameter at control, the restenosisrate in the right coronary artery group appearedhigher than that of the left anterior descending group,but the small number of patients with restenosis lim-its the statistical power of this comparison. This



FIGURE 2. Cumulative percentage of patients as a function of percent diameter stenosis as estimated with quantitative coronary angi-ography. Pre DCA Å before directional coronary atherectomy; Post DCA after directional coronary atherectomy; Control Å controlangiogram.

TABLE II Hemodynamic Data and Segmental Wall Motion in Patients Without Restenosis

Before Atherectomy

Right CoronaryArtery (n Å 23)

Left AnteriorDescending

Artery (n Å 50)

TotalPopulation(n Å 73)

Follow-Up Angiography



Artery (n Å 50)Total Population

(n Å 73)

Ejection fraction (%) 67 { 11 59 { 15§ 61 { 14 73 { 8‡ 62 { 17* 65 { 15‡

Fractional shortening (%)Segment 2 23 { 12 29 { 18 27 { 16 26 { 9 25 { 11 25 { 11Segment 3 41 { 19 41 { 16 41 { 17 48 { 16* 43 { 17 44 { 17Segment 4 45 { 18 38 { 19 41 { 19 51 { 14 42 { 19 45 { 18*Segment 5 39 { 16 26 { 19\ 30 { 19 43 { 13 30 { 18† 34 { 17‡

Segment 6 37 { 22 19 { 19Ø 25 { 22 42 { 19 27 { 22‡ 32 { 22‡

Segment 7 49 { 26 28 { 27\ 35 { 28 58 { 18* 39 { 27‡ 45 { 26‡

Segment 8 53 { 22 32 { 27Ø 39 { 28 62 { 17† 44 { 28‡ 50 { 26‡

Segment 9 56 { 21 35 { 26\ 42 { 26 61 { 19 46 { 28‡ 50 { 27‡

Segment 10 47 { 20 35 { 21§ 39 { 21 51 { 15 39 { 23 43 { 22Heart rate (beats/min) 70 { 9 70 { 12 70 { 11 71 { 10 68 { 13 69 { 12LV pressure (mm Hg)

End-diastolic 13 { 7 17 { 8 16 { 8 13 { 6 16 { 8 15 { 7Systolic 126 { 21 129 { 24 128 { 23 139 { 25* 141 { 29† 141 { 28‡

Mean aortic pressure (mm Hg) 93 { 14 94 { 16 94 { 16 95 { 14 99 { 18 98 { 17LV volume index (ml/m2)

End-diastolic 89 { 24 103 { 38 99 { 35 86 { 34 106 { 41 99 { 40End-systolic 30 { 12 45 { 31 40 { 27 27 { 20 45 { 39 39 { 34

*p õ0.05; †p õ0.01; ‡p õ0.001 before atherectomy versus at control in the same group; §p õ0.05; \p õ0.01; Øp õ0.001 right coronary artery group versus leftanterior descending artery group.

Values are expressed as mean { SD.LV Å left ventricular.

study confirms the importance of late loss whencompared with the acute gain that was expressed bya significant enlargement of the coronary lumen afteroptimal atherectomy (on average 1.84 mm). Moreprecisely, the loss index expressed as the late lossdivided by the initial gain was 0.49 { 0.42.

Evaluation of left ventricular function: The parame-ters of LV function were analyzed according to pres-ence or absence of angiographic restenosis as pre-viously defined. Main results are summarized inTables II and III with respect to the vessel dilatedand the restenosis process. Heart rate and mean aor-

tic pressure did not change significantly between the2 examinations. We observed only a mild to mod-erate elevation in LV systolic pressure in patientswithout restenosis at control (/12 { 23 mm Hg, põ0.001). Globally, there was no significant variationin LV volumes, but ejection fraction increased in thegroup without restenosis (/4 { 9%, p õ0.005). Incontrast, there was no significant variation in ejectionfraction (03 { 9%, p Å NS) in patients with angio-graphic restenosis. Conversely, regional wall motionanalysis demonstrated a highly significant improve-ment in segmental shortening, most of which was



TABLE III Hemodynamic Data and Segmental Wall Motion in Patients With Restenosis

Before Atherectomy



Artery (n Å 13)


Follow-Up Angiography



Artery (n Å 13)


Ejection fraction (%) 68 { 8 66 { 9 67 { 8 68 { 7 62 { 8 65 { 8Fractional shortening (%)

Segment 2 21 { 9 34 { 22 29 { 18 23 { 4 37 { 22 31 { 18Segment 3 42 { 21 43 { 20 43 { 20 43 { 15 47 { 22 45 { 19Segment 4 41 { 20 47 { 21 45 { 21 44 { 15 41 { 22 43 { 20Segment 5 31 { 16 33 { 24 33 { 21 36 { 17 26 { 21 30 { 20Segment 6 38 { 18 24 { 26 30 { 24 33 { 17 17 { 17 23 { 18Segment 7 53 { 20 32 { 32 40 { 30 50 { 14 27 { 21 36 { 22Segment 8 57 { 17 38 { 29 46 { 27 58 { 14 33 { 21 43 { 22Segment 9 58 { 16 43 { 28 49 { 24 59 { 15 33 { 18 44 { 21Segment 10 50 { 17 43 { 20 46 { 19 39 { 8 33 { 10 36 { 9*

Heart rate (beats/min) 61 { 7 63 { 10 62 { 9 63 { 9 67 { 8 65 { 9LV pressure (mm Hg)End-diastolic 15 { 6 15 { 8 15 { 7 14 { 4 13 { 8 13 { 6Systolic 122 { 13 124 { 25 123 { 21 131 { 17 134 { 25 133 { 22*

Mean aortic pressure (mm Hg) 91 { 8 97 { 20 95 { 16 98 { 11 100 { 21 99 { 18LV volume index (ml/m2)

End-diastolic 87 { 21 91 { 16 90 { 18 88 { 18 84 { 21 85 { 19End-systolic 28 { 11 32 { 10 30 { 10 28 { 11 31 { 7 30 { 9

*p õ0.05 before atherectomy versus at control in the same group.Values are expressed as mean { SD.LV Å left ventricular.

FIGURE 3. Segmental wall motion of the left ventricle in patients with left anterior descending lesion and without restenosis at controlangiogram. Left ventricular segments (Seg) are plotted on the X-axis and fractional shortening on the Y-axis. Normal range values areindicated as mean { 2 SD. Note the absence of significant variation in percent shortening of basal segments. Abbreviations as inFigure 2.

accounted for by anterior zones (/10% and /11%for segments 7 and 8, respectively, p õ0.001). Atbaseline examination, regional shortening was sig-nificantly abnormal, corresponding to a net hypoki-nesia due to myocardial stunning in patients withunstable angina, or to depressed contractility linkedto underlying necrosis (Figures 3 and 4). This featurewas far more pronounced in anterior than in inferiorsegments. In the restenosis group, 9 patients (41%)had prior myocardial infarction, and 5 patients hadinferior and 4 anterior infarction. The vascular dis-tribution was not statistically different when com-pared with that of the 37 patients (51%) with prior

infarction in the group without restenosis: 11 patientshad inferior and 26 anterior infarctions. Improve-ment in segmental wall motion was not influencedby previous infarction in the vascular territory con-cerned with atherectomy. Indeed, multivariate anal-ysis of variance for repeated measurements did notshow significant interaction between previous myo-cardial infarction and percent fractional shortening(for segment 7, F Å 2.02, p Å 0.15). In the absenceof restenosis, the time course of the regional wallmotion in patients with myocardial infarction did notexhibit differences when compared with that of theother patients. However, subgroup analysis demon-



FIGURE 4. Segmental wall motion of the left ventricle in patients with left anterior descending lesion and restenosis at control angio-gram. The left ventricular wall was considered to be hypokinetic if percent shortening of ¢2 contiguous segments (Seg) were õ2 SDvalues from the mean value. Abbreviations as in Figure 2.

strated that patients with myocardial infarction hadless pronounced improvement in percent fractionalshortening (/5 { 12% for segment 7, p õ0.05), re-flecting a persistent significant depression in LV con-tractility. Similarly, segmental motion improvementin inferior and basal regions at control was smallerin LV segments with previous myocardial infarction.Again, improvement in regional contractility was ob-served in the absence of restenosis only, and was farmore pronounced in anterior than inferior zones.

For the entire group with left anterior descendingartery lesions, we found a weak, although significant,negative correlation between percent diameter ste-nosis at control, and the variation in the percentageof segment shortening in the anterior wall (r Å00.28, p õ0.03; r Å 00.26, p õ0.01; r Å 00.26, põ0.02 for zones 7, 8, and 9, respectively).

DISCUSSIONAfter results of the large controlled trials,4,5 the

interest in DCA has been questioned.3 Nevertheless,the advantage of optimal atherectomy was clearlyindicated in retrospective studies that used thematching approach.16 All strategies that have aimedat an optimal increase in the vascular lumen havelimited the incidence of restenosis even though latedecrease in vessel lumen appears to be more pro-nounced using DCA.3,9,17 Our data on restenosisseems to be somewhat surprising since the popula-tion included a large proportion of patients with un-stable angina. However, our optimizing techniqueadded complementary balloon angioplasty to a greatnumber of cuts (on the average 20 to 30 per lesion)whose efficacy was demonstrated by the substantialamount of atheroma fragments generally collected.By using large devices and postdilation for most ar-teries, even for those as small as 3.0 mm in diameter,we obtained, on average after the procedure, a min-imum lumen diameter ú3.0 mm. These results arealmost identical to those found in the Optimal Ath-

erectomy Restenosis Study (3.15 mm).6 We mustemphasize that most of the vessels treated in our ex-perience were larger than those in previous reportedstudies. This may have favored a lower restenosisrate.17,18 Another explanation might involve thethrombotic material that is very frequently depositedon freshly ruptured plaques.19 Indeed, the residualthrombus that is left after angioplasty is consideredas a strong predictive parameter of restenosis.20 Byremoving a large amount of atherosclerotic material,we may have reduced thrombus incorporated in thevascular scar and limited intimal proliferation. Fi-nally, the population size was rather small comparedwith size in other controlled studies6,21 and the se-lected angiograms may have created systematic biasin our results.

Improvement in left ventricular regional wall motion:When myocardial perfusion is chronically altered,revascularization of dysfunctional areas has dem-onstrated its usefulness in preserving LV contractilefunction.10–12 Improvements in regional wall motionand global ejection fraction of the left ventricle aftercoronary bypass surgery are well recognized.22,23

These clinical observations can be explained by theconcept of ‘‘hibernating myocardium,’’ which couldbe due in part to chronic myocardial ischemia butalso to repeated episodes of acute ischemia that cul-minate in stunning of cardiac muscle.24,25 After cor-onary angioplasty, and particularly in the setting ofunstable angina, improvements in LV wall contrac-tility have been conclusively documented.11,14,25–27

Accordingly, we found in this prospective studya significant variation in LV wall motion that wasdetected from analysis of end-diastolic and end-sys-tolic ventricular contours. A more thorough studyusing a frame-by-frame approach on the overall car-diac cycle may have strengthened our conclusions.In previous studies, we followed midterm evolutionin patients treated with balloon angioplasty for leftanterior descending artery lesions and we found ben-



eficial effects on asynergy of the anterior wall andon asynchrony of the anterior segments during thecontraction phase. Other parameters reflecting re-gional dysfunction, such as shortening of contractionduration, were also significantly improved in the ab-sence of angiographic restenosis.12 In the presentstudy, wall motion in the myocardial zone revascu-larized with atherectomy improved in a significantmanner, whereas parameters of LV loading did notsignificantly change from baseline values. In fact, weobserved a tendency toward an increase in LV sys-tolic and mean aortic pressures. The improvementwas more marked for anterior segments, and lesspronounced for inferior and basal segments of theleft ventricle. As reported in studies dealing with cor-onary balloon angioplasty, angiographically visiblerestenosis limits recovery of regional contractility.28

The binary distinction of restenosis may appear ar-bitrary since recent studies in humans have foundthat attenuation of the coronary blood flow reserveexists for stenosis as low as 40%.7,8 Indeed, we founda significant correlation between the degree of ste-nosis and 6-month variations in percent fractionalshortening for all 95 patients. By contrast, eventhough there is a correlation between percent diam-eter stenosis and the consecutive decrease in coro-nary perfusion reserve, the predictive value of theangiographic severity of a given lesion remains un-certain.8 Accordingly, we cannot draw individualconclusions on the outcome of myocardial perfusionfrom our study.

LV dysfunction may recover spontaneously afterrepeated episodes of ischemia or after the acutephase of myocardial infarction. Some investigatorsfound no significant variations when ventricular vol-umes and segmental motion were measured after the10th day of infarction, but others have clearly shownthat recovery of LV function was possible after adelay of 5 weeks.29 Although we obtained no dataconcerning myocardial viability and cardiac metab-olism, improvement in segmental wall motion favorsthe persistence of a heterogeneous population ofmyocytes having different degrees of functional al-terations.23 Unfortunately, clinical follow-up onlywas available for patients who developed myocardialinfarction after DCA; thus, our data cannot be ex-tended to procedures associated with creatine kinaseelevation more than twofold normal.30

Study limitations: Patients were prospectively se-lected on the basis of angiographic criteria, and thegroup with previous myocardial infarction may havehad a reduction in the magnitude of variation in LVcontractility since improvement in regional wall mo-tion was significant but obviously smaller in this sub-set. However, the distribution of patients with myo-cardial infarction was not different according to thecoronary territory treated or the incidence of reste-nosis. Pathophysiologic consequences of coronaryblood flow changes on LV function are strongly in-fluenced by the dominance of the right coronary sys-tem, and complementary analysis performed usinganother radiologic projection may have reinforced

data demonstrating the benefit of revascularizationapparent in this study that focused on the left coro-nary system.

Acknowledgment: We gratefully acknowledge thetechnical assistance of Dominique Fernandez andFrancoise Noailles for preparing the manuscript, andGerard Gouverneur for his help with the statisticalanalysis.

1. Hinohara T, Robertson GC, Selmon MR, Vetter JW, Rowe MH, Braden LJ,McAuley BJ, Sheehan DJ, Simpson JB. Restenosis after directional coronaryatherectomy. J Am Coll Cardiol 1992;20:623–633.2. Fishman RF, Kuntz RE, Corrozza JP, Miller M, Senerchia CC, Schnitt SJ,Diver DJ, Safian RD, Baim DS. Long-term results of directional atherectomy:predictors of restenosis. J Am Coll Cardiol 1992;20:1101–1110.3. Umans VAWM, Robert A, Foley D, Wijns W, Haine E, de Feyter PJ, SerruysPW. Clinical, histologic and quantitative angiographic predictors of restenosisafter directional atherectomy: a multivariate analysis of the renarrowing processand late outcome. J Am Coll Cardiol 1994;23:49–58.4. Topol EJ, Leya F, Pinkerton CA, Whitlow PL, Hofling B, Simonton C,Masden RR, Serruys PW, Leon M, Williams DO, King SB III, Mark DB, IsnerJM, Holmes DR, Ellis SG, Lee KL, Keeler GP, Hinohara T, Califf RM. Acomparison of directional coronary atherectomy with coronary angioplasty inpatients with coronary artery disease. N Engl J Med 1993;329:221–227.5. Adelman AG, Cohen EA, Kimball BP, Bonan R, Ricci DR, Webb JG, Lar-amee L, Barbeau G, Traboulsi M, Corbett BN, Schwartz L, Logan AG. A com-parison of directional coronary atherectomy with balloon angioplasty for lesionsof the left anterior descending coronary artery. N Engl J Med 1993;329:228–234.6. Simonton CA, Leon MB, Kuntz, Popma JJ, Hinohara T, Bersin RM, YockPG, Wilson BH. Acute and late clinical and angiographic results of directionalatherectomy in the Optimal Atherectomy Study (OARS) (abstr). Circulation1996;92(suppl I):I–545.7. Wilson RF, Johnson MR, Marcus ML, Aylward PEG, Skorton DJ, Collins,White CW. The effect of coronary angioplasty on coronary flow reserve. Cir-culation 1988;77:873–885.8. Uren NG, Melin JA, De Bruyne B, Wijns W, Baudhuin T, Camici PG.Relation between myocardial blood flow and the severity of coronary arterystenosis. N Engl J Med 1994;330:1782–1788.9. Kuntz RE, Safian RD, Carrozza JP, Fisjman RF, Monsour M, Baim DS. Theimportance of acute luminal diameter in determining restenosis after coronaryatherectomy or stenting. Circulation 1992;86:1827–1835.10. Braunwald E, Kloner RA. The stunned myocardium: prolonged, postisch-emic ventricular dysfunction. Circulation 1982;66:1146–1149.11. de Feyter PJ, Suryapranata H, Serruys PW, Beat K, van der Brabd M,Hugenholtz PG. Effects of successful percutaneous transluminal coronary an-gioplasty on global and regional left ventricular function in unstable anginapectoris. Am J Cardiol 1987;60:993–997.12. Colle JP, Legoff G, Carfora A, Delarche N, Kilpatrick DD, Besse P. Long-term beneficial effects of PTCA on early segmental relaxation in disease of theleft anterior descending artery. Angiology 1988;39:466–478.13. Reiber JHC, van der Zweet PMJ, von Land CD, Koning G, Loois G, ZornI, van der Brand M, Gerbrands JJ. On-line quantification of coronary angiogramswith the DCI system. Medicamundi 1989;34:89–98.14. Colle JP, Legoff G, Ohayon J, Bonnet J, Bricaud H, Besse P. Quantitativeframe by frame analysis of regional contraction and lengthening on left ven-tricular cineangiograms: application to the study of normal left ventricles andleft ventricles with mitral valve prolapse. Clin Cardiol 1986;9:43–51.15. Ingels NB, Daughters GT, Stinson EB, Alderman EL. Evaluation of meth-ods for quantitating left ventricular segmental wall motion in man using myo-cardial markers as a standard. Circulation 1980;61:966–972.16. Umans VAWM, Keane D, Foley D, Boersma E, Melkert R, Serruys PW.Optimal use of directional coronary atherectomy is required to ensure long-termangiographic benefit: a study with matched procedural outcome after atherec-tomy and angioplasty. J Am Coll Cardiol 1994;24:1652–1659.17. Kuntz RE, Hinohara T, Robertson GC, Safian RD, Simpson JB, Baim DS.Influence of vessel selection on the observed restenosis rate after endoluminalstenting or directional atherectomy. Am J Cardiol 1992;70:1101–1108.18. Foley DP, Melkert R, Serruys PW. Influence of coronary vessel size onrenarrowing process and late angiographic outcome after successful balloonangioplasty. Circulation 1994;90:1239–1251.19. Arbustini E, De Servi S, Bramucci E, Porcu E, Costante AM, Grasso M,Diegoli M, Fasani R, Morbini P, Angoli L, Boscarini M, Repetto S, Danzi G,Niccoli L, Campolo L, Lucreziotti S, Specchia G. Comparison of coronarylesions obtained by directional coronary atherectomy in unstable angina, stableangina, and restenosis after either atherectomy or angioplasty. Am J Cardiol1995;75:675–682.



20. Bauters C, Lablanche JM, MacFadden EP, Hamon M, Bertrand ME. Re-lation of coronary angioscopic findings at coronary angioplasty to angiographicrestenosis. Circulation 1995;92:2473–2479.21. Baim DS, Kuntz R, Sharma SK, Fortuna R, Feldman R, Senerchia C, DeFeoT, Popma JJ, Ho KKL, for the BOAT investigators. Acute results of the ran-domized phase of the balloon versus optimal atherectomy trial (BOAT) (abstr).Circulation 1995;92(suppl I):I–544.22. Vanoverschelde JLJ, Wijns W, Depre C, Essamri B, HeyndrickxGR, Borgers M, Bol A, Melin JA. Mechanisms of chronic regional post-ischemic dysfunction in humans. New insights from the study of nonin-farcted collateral-dependent myocardium. Circulation 1993;87:1513 –1523.23. Marwick TH, MacIntyre WJ, Lafont A, Nemec JJ, Salcedo EE. Metabolicresponses of hibernating and infarcted myocardium to revascularization. A fol-low-up study of regional perfusion, function, and metabolism. Circulation1992;85:1347–1353.24. Shen YT, Vatner SF. Mechanism of impaired myocardial function duringprogressive coronary stenosis in conscious pigs. Hibernation versus stunning?Circ Res 1995;76:479–488.25. Nienaber CA, Brunken RC, Sherman CT, Yeatman LA, Gambhir SS, Kri-vokapich J, Demer LL, Ratib O, Child JS, Phelps ME, Schelbert HR. Metabolic

and functional recovery of ischemic human myocardium after angioplasty. JAm Coll Cardiol 1991;18:966–978.26. Rosing DR, Cannon RO III, Watson RM, Bonow RO, Mincemoyer R, EwelsC, Leon M, Lakatos E, Epstein SE, Kent KM. Three-year anatomic, functionaland clinical follow-up after successful percutaneous transluminal coronary an-gioplasty. J Am Coll Cardiol 1987;9:1–7.27. van den Berg EK, Popma JJ, Dehmer GJ, Snow FR, Lewis SA, VetrotecGW, Nixon JV. Reversible segmental left ventricular dysfunction after coronaryangioplasty. Circulation 1990;81:1210–1216.28. Melgares R, Prieto JA, Azputarte J. Significant coronary restenosis limitsthe recovery of regional left myocardial dysfunction achieved after successfulcoronary angioplasty. Eur Heart J 1993;14:866–875.29. Galli M, Marcassa C, Bolli R, Giannuzzi P, Temporelli PL, Imparato A,Orrego S, Giubbini R, Giordano A, Tavazzi L. Spontaneous delayed recoveryof perfusion and contraction after the first 5 weeks after anterior myocardialinfarction. Evidence for the presence of hibernating myocardium in the infarctedarea. Circulation 1994;90:1386–1397.30. Kugelmass AD, Cohen DJ, Moscucci M, Piana RN, Senerchia C, KuntzRE, Baim DS. Elevation of the creatine kinase myocardial isoform followingotherwise successful directional coronary atherectomy and stenting. Am J Car-diol 1994;74:748–754.

impact of restenosis after optimal directional coronary atherectomy on regional left ventricular...

Documents