Quantitative two-dimensional echocardiographic measurements are major predictors of adverse cardiovascular events after acute myocardial infarction. The protective effects of captopril

Lamas, M Klein, B Sussex and S GoldmanM St. John Sutton, MA Pfeffer, T Plappert, JL Rouleau, LA Moye, GR Dagenais, GA

The protective effects of captoprilpredictors of adverse cardiovascular events after acute myocardial infarction.

Quantitative two-dimensional echocardiographic measurements are major

ISSN: 1524-4539 Copyright © 1994 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online

72514Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX

1994, 89:68-75Circulation

http://circ.ahajournals.org/content/89/1/68located on the World Wide Web at:

The online version of this article, along with updated information and services, is

http://www.lww.com/reprintsReprints: Information about reprints can be found online at [email protected]. E-mail:

Fax:Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters http://circ.ahajournals.org//subscriptions/Subscriptions: Information about subscribing to Circulation is online at

by guest on July 18, 2011http://circ.ahajournals.org/Downloaded from

http://circ.ahajournals.org/

68

Quantitative Two-dimensionalEchocardiographic Measurements Are Major

Predictors of Adverse Cardiovascular EventsAfter Acute Myocardial Infarction

The Protective Effects of Captopril

Martin St. John Sutton, FRCP; Marc A. Pfeffer, MD, PhD; Ted Plappert, CVT;Jean-Lucien Rouleau, MD; Lemuel A. Moye, MD, PhD; Gilles R. Dagenais, MD;

Gervasio A. Lamas, MD; Marc Klein, MD; Bruce Sussex, MD; Steven Goldman, MD;Francis J. Menapace, Jr, MD; John 0. Parker, MD; Sandra Lewis, MD;

Francois Sestier, MD, PhD; David F. Gordon, MD; Patricia McEwan, MD;Victoria Bernstein, MD; Eugene Braunwald, MD, for the SAVE Investigators,

Brigham and Women's Hospital, Harvard Medical School, Boston

Background Left ventricular enlargement after myocardialinfarction increases the likelihood of an adverse outcome. Inan echocardiographic substudy of the Survival and VentricularEnlargement (SAVE) Trial, we assessed whether captoprilwould attenuate progressive left ventricular enlargement inpatients with left ventricular dysfunction after acute myocar-dial infarction and, if so, whether this would be associated withimproved clinical outcome.Methods and Results Two-dimensional transthoracic

echocardiograms were obtained in 512 patients at a mean of11.1+3.2 days after infarction and were repeated at 1 year in420 survivors. Left ventricular size was assessed as left ventric-ular cavity areas at end diastole and end systole and leftventricular function as percent change in cavity area from enddiastole to end systole. Patients were randomly assigned toplacebo or captopril, and the incidence of adverse cardiovas-cular events consisting of cardiovascular death, heart failurerequiring either hospitalization or open-label angiotensin-converting enzyme inhibitor therapy, and recurrent infarctionwere determined over a follow-up period averaging 3.0±0.6years. Irrespective of treatment assignment, baseline left ven-tricular systolic area and percent change in area were strongpredictors of cardiovascular mortality and adverse cardiovas-cular events. At 1 year, left ventricular end-diastolic and

T he prognosis in survivors of acute myocardialinfarction depends on multiple factors, whichrelate both to the acute event, such as infarct

size, location, and transmurality, as well as to charac-teristics preceding the infarction such as age, sex, riskfactors, and prior infarction.'-8 However, the single most

Received April 1, 1993; revision accepted July 20, 1993.From Brompton Hospital, London, England (M.S.J.S.); Har-

vard Medical School and Brigham and Women's Hospital,Boston, Mass (M.A.P., T.P., G.A.L., E.B.); Universite de Sher-brooke, Quebec (J.L.R.); University of Texas Health ScienceCenter at Houston (L.A.M.); Universite de Montreal (G.R.D.);H6pital du Sacre-Coeur de Montreal (M.K.); Memorial Univer-sity of St. John's Newfoundland (B.S.); VA Medical Center,Tucson, Ariz (S.G.); Geisinger Medical Center, Danville, Pa

end-systolic areas were larger in the placebo than in thecaptopril group (P=.038, P=.015, respectively), and percentchange in cavity area was greater in the captopril group(P=.005). One hundred eleven of the 420 l-year survivors withl-year echo measurements (26.4%) experienced a major ad-verse cardiovascular event, and these patients had more than athreefold greater increase in left ventricular cavity areas thanthose with an uncomplicated course. Sixty-nine patients withadverse cardiovascular events were in the placebo groupcompared with 42 patients in the captopril-treated group (arisk reduction of 35%, P=.010).

Conclusions Two-dimensional echocardiography providesimportant and independent prognostic information in patientsafter infarction. Left ventricular enlargement and functionafter infarction are associated with the development of adversecardiac events. Attenuation of ventricular enlargement withcaptopril in these patients was associated with a reduction inadverse events. This study demonstrates the linkage betweenattenuation of left ventricular enlargement by captopril afterinfarction and improved clinical outcome. (Circulation.1994;89:68-75.)Key Words * left ventricle * myocardial infarction -

angiotensin * enzymes * heart failure, congestive

important determinant of survival is postinfarction leftventricular function.24 During the last decade, a num-ber of new therapeutic strategies have been developedspecifically to attempt to reduce infarct size and therebyminimize the extent of left ventricular dysfunction.6-8These have consisted of either restoring flow to the

(F.J.M.); Kingston General Hospital, Kingston, Ontario(J.O.P.); Oregon Heart Institute, Portland, Ore (S.L.); H6pitalNotre-Dame de Montreal, Quebec (F.S.); Cardiology Associ-ates, Des Moines, Iowa (D.G.); The Wellesley Hospital, Tor-onto, Ontario (P.McE.); University of British Columbia, Van-couver (V.B.).

Correspondence to Marc A. Pfeffer, MD, PhD, CardiovascularDivision, Brigham and Women's Hospital, 75 Francis St, Boston,MA 02115.



St. John Sutton et al Echocardiographic Effects of Captopril: The SAVE Study 69

infarct-related coronary artery9J0 or reducing left ven-tricular loading conditions.11-17 There is great need for asimple, reliable method of assessing left ventricular sizeand function early after infarction, which can be usedduring long-term follow-up to assess prognosis, selectfuture therapy, and determine the efficacy of newmechanical and pharmacological interventions. Al-though two-dimensional echocardiography enables di-rect visualization of the entire left ventricle in orthogo-nal tomographic images, most postinfarction trials haveused radionuclide ventriculography5,18 because of tech-nical difficulties in obtaining high-quality two-dimen-sional echocardiograms in multiple views. We usedtwo-dimensional echocardiography to define prospec-tively left ventricular size from multiple orthogonalviews in patients randomized in the Survival and Ven-tricular Enlargement (SAVE) Trial and related it toclinical outcome.The SAVE Trial demonstrated that captopril therapy

improved survival and reduced the number of adversecardiovascular events in patients who survived acutemyocardial infarction with evidence of left ventriculardysfunction.19 The goals of this prospective two-dimen-sional echocardiographic study were to determinewhether the results of quantitative analysis of two-dimensional echocardiograms obtained early after in-farction could be used to (1) predict mortality andadverse cardiovascular events, (2) reliably detect pro-gressive left ventricular enlargement, (3) determinewhether left ventricular enlargement at baseline andprogressive enlargement over the subsequent year wasassociated with excess mortality and adverse cardiovas-cular events, (4) determine whether the improved clin-ical outcome resulting from captopril therapy was asso-ciated with attenuation of the time-dependent processof left ventricular enlargement.

MethodsDuring the enrollment phase of the SAVE Trial, 24 of the 45

centers participated in the two-dimensional echocardiographicancillary study. Two-dimensional echocardiograms of theshort axis of the left ventricle were obtained from the paraster-nal region at three levels: the mitral valve, the high, and themidpapillary muscle levels. The long axis of the left ventriclewas imaged from the apex in the four-chamber view and theapical long-axis or the two-chamber view.

All two-dimensional echocardiograms were submitted to theechocardiographic core laboratory at the Brigham and Wom-en's Hospital for assessment of technical quality and todetermine whether they were adequate for quantitative analy-sis. The definition of a technically acceptable two-dimensionalechocardiogram was one that included images of the long andshort axes of the left ventricle in a minimum of three of fiveviews. These two-dimensional echocardiographic images weretransferred to the hard disk of a Freeland medical off-linecomputer analysis system and digitized to obtain cavity areasat end diastole and end systole. In patients whose echocardio-grams were found to be suitable for analysis, repeat studieswere required at their 1-year visit.

Left ventricular short-axis areas were summed, and theaverage short-axis areas for end diastole and end systole werecalculated. Similarly, average left ventricular long-axis areaswere calculated for end diastole and end systole. Left ventric-ular size was assessed as the sum of the average short-axis andthe average long-axis areas at end diastole and end systole.Thus, we defined left ventricular size from multiple orthogonaltwo-dimensional echocardiographic images at baseline and at

1 year. Percent change in left ventricular area was calculatedas the difference between end-diastolic area and end-systolicarea expressed as a percentage of end-diastolic area and wasused as an index of left ventricular contractile function. Theprognostic significance of end-diastolic and end-systolic leftventricular areas and percent change in left ventricular area atbaseline was assessed by relating them to the clinical endpoints at long-term follow-up. The end points that describe theclinical outcome were defined prospectively as cardiovasculardeath, development of heart failure requiring hospitalizationor open-label angiotensin-converting enzyme inhibition ther-apy, and recurrent myocardial infarction.19

Left ventricular end-diastolic and end-systolic areas andpercent change in left ventricular area at baseline, and thechanges from baseline to 1 year were compared in the twotreatment groups to assess whether captopril therapy alteredleft ventricular enlargement and/or left ventricular functionafter infarction. The clinical outcomes in the placebo and thecaptopril treatment groups were compared in terms of thenumber of adverse cardiovascular events (cardiovasculardeath, heart failure requiring either hospitalization or open-label captopril, and recurrent myocardial infarction). A rela-tion was sought between left ventricular enlargement andpercent change in left ventricular area from baseline to 1 yearand the incidence of adverse cardiovascular events irrespectiveof treatment assignment. In addition, we assessed whether thebeneficial effects of captopril therapy on clinical outcome wererelated to changes in left ventricular areas or to changes in leftventricular function expressed as percent change in area.

Statistical Methods: ReproducibilityThe reproducibility of the two-dimensional echocardio-

graphic measurements of left ventricular end-diastolic andend-systolic areas and percent change in area was assessed byone observer who digitized the two-dimensional echocardio-grams of 50 patients three times but was blinded to the resultsof the previous analyses. These three independent assessmentsof left ventricular areas and percent change in area were usedto provide information on intrasubject variability. The intra-subject variability was pooled, resulting in an estimate of thevariance of reproducibility. The individual mean differencesand the standard deviations of the difference of reproducibilitywere calculated as previously described20 and are shown inTable 1.Assessment of the relation between the echocardiographic

measurements at baseline, the changes from baseline to 1 year,and adverse cardiovascular events was examined by bothunivariate and multivariate analyses. The univariate analysisconsisted of the patients' results being stratified into quartilesbased on left ventricular end-diastolic and end-systolic areasand percent change in area. For each echocardiographicvariable, the relation between the number of clinical eventswas tabulated in each of these quartiles. X 2 testing was used toassess statistical significance, with a value of P<.05 requiredfor statistical significance.

Multivariate analysis was used to determine whether leftventricular end-diastolic and end-systolic areas and percentchange in area at baseline provided predictive informationindependent of other patient characteristics for total mortal-ity. A model to predict survival was constructed by Coxproportional hazard analysis using the SAVE clinical database. The model included independent variables known toinfluence survival: age, sex, prior myocardial infarction, diabe-tes mellitus, hypertension, use of thrombolysis, and radionu-clide ejection fraction. The dependent variable was the time todeath. Left ventricular end-diastolic and end-systolic areas andpercent change in area were added individually to the model toestablish whether they improved the predictive accuracy of themodel. Relative risks were computed for a change of 1.96times the standard deviation of reproducibility for each echo-cardiographic variable. For each model, a P value was gener-



70 Circulation Vol 89, No 1 January 1994

TABS 1. Reproducibility Analysis of the Echo Data inthe SAVE Trial

StandardEcho Measurement Mean DeviationDiastolic cavity area, cm2 72.037 3.472

Systolic cavity area, cm2 51.477 2.933

Change in area, % 0.291 0.029

ated for each independent variable tested. This value is anassessment of the relation between that independent variableand total mortality after adjusting for the relation between theeffect on total mortality of the other independent variables inthe model. A value of P<.05 was required for statisticalsignificance.

ResultsThe two-dimensional echocardiograms from 785 pa-

tients were submitted to the echocardiographic corelaboratory from 24 centers that attempted to obtainechocardiograms in all patients randomized; 512 (65%)baseline studies were accepted for quantitative analysis.These were obtained at a mean of 11.1±3.2 days afterthe index myocardial infarction. Thirteen patients diedduring the follow-up period from noncardiovascularcauses, leaving a cohort of 499 patients. There were 55deaths within the first year before the follow-upechocardiogram (24 in the placebo and 31 in the cap-topril group, P=.275). Four hundred twenty of thesesurvivors (95%) had repeat two-dimensional echocar-diograms that were judged acceptable for quantitativeanalysis at 1 year.The clinical features of the patients in the echocar-

diographic substudy resembled those in the overalltrial.19 Furthermore, patients randomly assigned to pla-cebo or to captopril within the echocardiographic sub-study were similar except for the peak creatine phos-phokinase levels, which were significantly higher in thecaptopril-treated group (Table 2). Patients were fol-lowed up from 2.0 to 4.3 years (mean of 3.0 years) afterrandomization.

Baseline left ventricular end-diastolic and end-sys-tolic areas and percent change in area were all strong,independent predictors of cardiovascular mortality (Ta-ble 3 and Fig 1). These baseline echocardiographicvariables were added individually to the multivariatemodel for survival, which included factors well known toinfluence survival (age, sex, diabetes, hypertension,prior infarction, thrombolysis, and radionuclide ejectionfraction). When added to this model, left ventricularend-systolic area and percent change in left ventriculararea emerged as some of the strongest independentpredictors of clinical outcome (Tables 4 and 5).

TABLE 2. Baseline Characteristics of the Echo CohortIn the SAVE Trial

Placebo CaptoprilCharacteristic (n=259) (n=253)Mean age, y 58.3±11.7 58.9+10.8Male/female 206/53 211/42Mean RVG-EF 31.2±6.7% 30.7±6.6%Mean peak CPK(as % of normal) 1295±1246 1533±1300*Prior myocardial infarction 92 (35.5%) 76 (30.0%)History of hypertension 84 (32.4%) 80 (31.6%)History of diabetes 56 (21.6%) 44 (17.4%)Thrombolysis 95 (36.7%) 113 (44.7%)PTCA 33 (12.7%) 50 (19.8%)

RVG-EF indicates radionuclide ventriculographic ejection frac-tion; CPK, creatine phosphokinase; and PTCA, percutaneoustransiuminal coronary angioplasty.

*P<.05.

Left ventricular end-diastolic and end-systolic areasand percent change in left ventricular area at baselinewere similar in the captopril and the placebo treatmentgroups (Table 6). At 1 year, left ventricular end-diastol-ic and end-systolic areas had increased in both treat-ment groups, but the increases in areas were signifi-cantly greater in the patients treated with placebo thanin those treated with captopril (P=.023 for diastole,P=.021 for systole) (Fig 2). This greater incrementresulted in larger left ventricular areas in the 1-yearsurvivors in the placebo-treated group (P=.038 andP=.015, respectively) (Fig 2). Percent change in leftventricular area was significantly greater at 1 year in thecaptopril-treated group compared with placebo(29+8% versus 27±7%, P=.005), although comparisonof the changes over 1 year within the respective treat-ment groups was not statistically significant (-0.1+7%versus -1.3+7%, P=.105).One hundred eleven of the 420 1-year survivors with

1-year echo measurements (26.4%) sustained at leastone major cardiovascular event (defined as cardiovascu-lar death, heart failure requiring either hospitalizationor open-label captopril therapy, or recurrent infarction)during late follow-up. The event rate was higher in theplacebo treatment group (31.8%) compared with thecaptopril-treate.l group (20.7%) (Fig 3 and Table 7),indicating a risk reduction of 35% (P=.010).

Patients who sustained adverse cardiovascular eventsirrespective of treatment assignment had more thanthreefold greater increases in left ventricular end-dia-stolic and end-systolic areas from baseline to 1 year

TABLE 3. Influence of Baseline Echo Variables in Predicting Mortality in the SAVE Trial

Univariate Mutvariate

Relative RelatfveVariable P Risk P RiskDiastolic cavity area <.001 1.25 .029 1.11

Systolic cavity area <.001 1.25 .003 1.12Change in area <.001 2.06 <.001 1.55



St. John Sutton et al Echocardiographic Effects of Captopril: The SAVE Study

60

i 50

.- 40

1-

30

Q) 20

a-

10

10

11 Ill lV

60

R 50

_ 40

h-0 30:,-

Q 20

10

0

1 11 Ill lV

60

%&50

_ 40b-0 30

4' 20

10

0

Left Ventricular Size at Baseline in Quartiles Percentage Changein Cavity Area atBaseline Quartiles

FIG 1. Graphs show relations between left ventricular end-diastolic area (left), end-systolic area (middle), and percent change in cavityarea (right) at baseline and subsequent cardiovascular mortality. The study population of 512 patients is divided into quartiles ofincreasing left ventricular areas (I through IV) and demonstrates that an increase in area is associated with increase in cardiovascularmortality. Percent change in quartiles (increasing from I through IV) demonstrates that decreasing percent change in area is stronglyassociated with increasing cardiovascular mortality.

compared with patients who experienced no adversecardiovascular events (5.35+±9.88 versus 1.48±9.01 cm2

in diastole, P<.001, and 6.37±10.29 versus 1.35±9.62cm2 in systole, P<.001) (Fig 4). Furthermore, the inter-val change in percent change in left ventricular area at1 year was significantly lower in patients who sustainedadverse cardiovascular events compared with patientswho did not (-3±6% versus 0±7%, P<.001). Captopriltherapy resulted in a trend toward less deterioration inthe percent change in area (-1.3 +7.0 versus -0.1±7.0,P=.105; placebo versus captopril). Among patients whodid not sustain adverse cardiovascular events, the in-crease in left ventricular area was much less in thecaptopril-treated group both in diastole (0.38±9.21versus 2.67±8.67 cm2, P=.02) and in systole (0.26±9.98versus 2.53±9.11 cm2, P=.038). Although captoprilreduced the number of patients with cardiovascularevents, the changes in left ventricular cavity areas in thecaptopril-treated patients with cardiovascular events

were similar to the patients treated with placebo whoexperienced adverse cardiovascular events (Fig 5).

DiscussionThe prognosis of patients surviving acute myocardial

infarction is influenced by the interaction of a largenumber of factors including age, sex, the presence ofrisk factors such as hypertension and diabetes, previousinfarction, residual or recurrent ischemia, the extent ofcoronary artery disease, and the presence of ventriculararrhythmias. However, the most powerful determinantsare left ventricular function24 and left ventricular size.5Furthermore, left ventricular enlargement after myocar-dial infarction is associated with decreased survivalfrom congestive heart failure,l-3 the risk of death in-creasing in direct relation to left ventricular size.21

Quantitative two-dimensional echocardiography hasprovided important insights into the mechanism andtemporal sequence of left ventricular enlargement and

TABLE 4. Multivariate Regression Model for Predicting Total Mortality Systolic Cavity Area

Confidence Interval

Relative Lower Upper WaldRisk Bound Bound X2 P

Age 1.52 1.23 1.89 14.72 <.001

Sex 0.93 0.54 1.60 0.08 .78

Prior myocardial infarction 1.57 1.05 2.37 4.78 .028

Diabetes 1.19 0.75 1.87 0.55 .46

Hypertension 1.33 0.89 2.00 1.92 .166

Thrombolysis 0.70 0.43 1.13 2.17 .141

LVEF 0.57 0.42 0.77 12.92 <.001

Systolic area 1.12 1.04 1.21 9.06 .003

The relative risk is computed for a change in 10 units for age and ejection fraction. It is computed for a changeof 1.96 times the standard deviation of reproducibility for the echo measures (5.75 for systolic cavity area). LVEFindicates left ventricular ejection fraction.

71




TABLE 5. Multivariate Regression Model for Predicting Total Mortality Percent Change in Area

Confidence Interval

Relative Lower Upper WaldRisk Bound Bound X2 P

Age 1.50 1.21 1.85 13.52 <.001

Sex 1.60 0.95 2.71 0.03 .86

Prior myocardial infarction 1.56 1.04 2.34 4.66 .031

Diabetes 1.17 0.74 1.84 0.45 .504

Hypertension 1.43 0.95 2.14 2.97 .085

Thrombolysis 0.71 0.44 1.15 1.93 .165

LVEF 0.62 0.46 0.84 9.47 .002

% Change 0.65 0.52 0.79 17.33 <.001

The relative risk is computed for a change in 10 units for age and ejection fraction. It is computed for a changeof 1.96 times the standard deviation of reproducibility for the echo measures (0.057 for percent change in area).LVEF indicates left ventricular ejection fraction.

contractile dysfunction after infarction.22-27 The extentof abnormal left ventricular wall motion by two-dimen-sional echocardiography has correlated with infarct sizeassessed as the peak level of cardiac enzyme release andalso with incidence of late complications, although notwith survival.28'29 Longitudinal two-dimensional echo-cardiographic studies have demonstrated that left ven-tricular enlargement is a complex and dynamic processwith at least two components. The early phase occurswithin the first 72 hours of acute infarction and resultsfrom stretching of the infarct zone due to slippage of thenecrotic myofibrils22'30 before the tensile strength of theinfarcted tissue has been increased by collagen deposi-tion.31 The second phase, which involves the entireventricle, has a more protracted time course and maycontinue for months or years and consists of dilationand hypertrophy of the noninfarct zone in response tothe loss of contractile elements and increased wallstress. This dynamic process of changing left ventricularshape and architecture in response to the loss of myo-cytes has been termed left ventricular remodeling.3' Anumber of two-dimensional echocardiographic studieshave clearly shown that appropriately timed interven-tions with thrombolysis and perturbations in left ven-tricular loading conditions with pharmacological agentscan favorably influence left ventricular remodeling andcontractile function late after infarction.'2-17'3233We prospectively defined left ventricular size at end

diastole and end systole from multiple orthogonal echo-cardiographic images and calculated percent change inleft ventricular area as an index of contractile function.These echocardiographic measurements were used to

TABLE 6. Echo Two-dimensional Variables at Baseline

Placebo Captopril(n=259) (n=253)

LV diastolic area, cm2 71.6+12.8 70.7±+12.3LV diastolic area index, cm2 38.1±6.5 37.4±6.0

LV systolic area, cm2 52.2±12.6 50.7±12.0

LV systolic area index, cm2 27.7+6.6 26.8±5.9

% Change in LV area 27.8±6.5 28.8±5.8

LV indicates left ventricular.

quantitate left ventricular size and function at baselineand at 1 year to characterize the changes that occurredafter infarction. Quantitating left ventricular size interms of areas in these deformed ventricles avoids usingalgorithms that assume a specific or fixed geometricshape. We related these echocardiographic variables atbaseline and the interval changes from baseline to 1year to clinical outcome to determine whether thetwo-dimensional echocardiograms added prognostic in-formation independent of the baseline clinical patientdescriptors and whether the beneficial effects of capto-pril resulted from modification of the time-dependentprocess of left ventricular enlargement.

Left ventricular areas at end diastole and at endsystole and percent change in area at baseline corre-lated strongly with subsequent cardiovascular mortality

80

V'io

dS'U

tD E

75

70

65

60

eCD)

c-J

v 5

Lg 4

0

Placebo .. K O_~~~~~~-e 3

p<O.038 a X.mCaptopril j0 2

Baseline 1 Year

40 'Baseline 1 Year

p- O.o23

Captopril Placebo

C',

U20_1

co

Captoprll Plcebo

FIG 2. Graphs of left ventricular end-diastolic (top) and end-systolic (bottom) areas at baseline and at 1 year in the twotreatment groups show that captopril attenuates left ventricularenlargement.




Va'Uto -,

(1 c

6)-0

40

30

20

10

0Placebo Captopril(69/217) (42/203)

FIG 3. Bar graph shows the incidence of adverse cardiovascu-lar events (cardiovascular death, heart failure requiring hospital-ization or open-label angiotensin-converting enzyme inhibitortherapy, and recurrent myocardial infarction) in the placebo-treated and captopril-treated patients during a mean follow-upperiod of 3 years.

(Fig 1). The prognostic value of left ventricular areas isconsistent with previous reports obtained from radionu-clide and contrast angiocardiography that ventricularsize (volumes or areas) are strong independent predic-tors of survival after myocardial infarction.5 When leftventricular end-systolic area and percent change incavity area were added to a multivariate model predict-ing survival, which included factors well known toinfluence outcome (age, sex, prior infarction, hyperten-sion, diabetes, use of thrombolysis, and radionuclideejection fraction), they emerged as two of the mostpowerful independent predictors of clinical outcome;percent change in area was even greater than ejectionfraction, which hitherto had been widely held to be thestrongest noninvasively determined predictor of out-come after infarction. These findings indicate that quan-titative two-dimensional echocardiography early afterinfarction provides important prognostic informationthat is independent of the conventional patientdescriptors.The effects of angiotensin-converting enzyme inhibi-

tion with captopril on left ventricular size after infarc-tion were clearly demonstrated when the intervalchanges in the echocardiographic end-diastolic andend-systolic areas from baseline to 1 year were com-pared in the captopril and the placebo treatmentgroups. Left ventricular end-diastolic and end-systolicareas were similar at baseline in the two treatment

8

a)Lj -E

LQ)t°EU

._c

-e

6

4

Diastole

- ~ --rp <<O.OOlj-

2 -

0No

CV Events(n =309)

8

6

4

2

0AdverseCV Events

(n= 11 1)

TABLE 7. Clinical End Points by Treatment GroupAmong 420 Echo Patients With 1-Year Echo Data

Placebo CaptoprilEnd Point (n=217) (n=203)Cardiovascular death 22 (10.1%) 12 (5.9%)

CHF requiring ACE 36 (16.6%) 12 (5.9%)

Hospitalization CHF 41 (18.9%) 18 (8.9%)

Severe heart failure (either CHF,ACE, or hospitalization CHF) 54 (24.9%) 26 (12.8%)

Recurrent myocardial infarction 35 (16.1%) 18 (8.9%)

Number of patients with any ofthe above 69 (31.8%) 42 (20.7%)CHF indicates congestive heart failure; ACE, angiotensin-

converting enzyme.

groups, and at 1 year left ventricular enlargement hadoccurred in both groups but significantly less so in thecaptopril-treated patients. This finding is consistentwith previous reports that captopril attenuates leftventricular enlargement after infarction.14-'7 By con-trast, there was no significant difference in the intervalchange in percent change in cavity area at 1 year in thecaptopril-treated compared with the placebo-treatedpatients.

In the parent SAVE Trial, captopril resulted inimproved clinical outcome.19 Similarly, in this echocar-diographic substudy, captopril significantly reduced theincidence of adverse cardiovascular events defined ascardiovascular death, heart failure requiring hospital-ization or open-label captopril, and recurrent myocar-dial infarction by more than one third (35%) comparedwith placebo. Paired two-dimensional echocardio-graphic assessment of left ventricular size at baselineand 1 year provided insight into the possible etiologicalfactors involved in the development of adverse cardio-vascular events. Irrespective of treatment assignment,the increase in left ventricular areas at 1 year in patientsexperiencing adverse cardiovascular events was three-fold greater than in patients with a benign course afterinfarction. These two-dimensional echocardiographicfindings indicate that left ventricular enlargement afterinfarction is associated with an increased incidence ofadverse cardiovascular events.

Systole

NoCV Events

(n =309)

FIG 4. Bar graphs show changesin left ventricular area betweenbaseline and 1 year in patientswho experienced adverse cardio-vascular (CV) events comparedwith those with an uncomplicatedclinical course after infarction.

AdverseCV Events

(n=111)




cGD

t°E

_D O_

C-

.U C-

L..

G_D-~ m

8

6

4

2

0

Diastole8

6

4

2

0No Adverse

CV Events CV Events(n-148) (n-161) (n-69) (n-42)

To determine whether attenuation of left ventricularenlargement with captopril translated into improvedclinical outcome, we compared the interval changes inend-diastolic and end-systolic areas from baseline to 1year in patients who sustained adverse cardiovascularevents with those without adverse events both in thecaptopril and placebo treatment groups. Captopril ther-apy resulted in a significant reduction in the number ofpatients who exhibited left ventricular enlargement andhad a poor clinical outcome. Even in patients who hadno adverse cardiovascular events, captopril therapy re-sulted in attenuation of left ventricular enlargement,suggesting that echocardiographically detected deterio-ration was apparent in the clinically well patients onplacebo. Thus, captopril therapy reduced the size of thereservoir of patients at high risk of sustaining majoradverse cardiovascular events.We conclude that quantitative two-dimensional echo-

cardiography in survivors of acute myocardial infarctionprovides powerful and independent prognostic informa-tion during long-term follow-up. Progressive left ven-

tricular enlargement can be detected reliably by two-dimensional echocardiography and, when present, isassociated with the development of adverse cardiovas-cular events. Although attenuation of left ventricularenlargement with captopril has been previously re-ported, our study had sufficient sample size and dura-tion to provide the critical linkage between the struc-tural changes in the left ventricle and clinical eventsafter myocardial infarction. This finding may explain inpart the beneficial effects on clinical outcome in theparent SAVE Trial.

AcknowledgmentsThe SAVE Study was funded by a grant from the Bristol-

Myers Squibb Institute for Pharmaceutical Research and by agrant from the Medical Research Council of Canada.

References1. Hammermeister KE, DeRouen TA, Dodge HT. Variables pre-

dictive of survival in patients with coronary disease: selection byunivariate and multivariate analyses from the clinical, electrocar-diographic, exercise, arteriographic, and quantitative angiographicevaluations. Circulation. 1979;59:421-430.

2. Multicenter Postinfarction Research Group. Risk stratificationand survival after myocardial infarction. N Engl J Med. 1983;309:331-336.

FiG 5. Bar graphs show changesin left ventricular area betweenbaseline and 1 year in patients whodid versus those who did not expe-rience an adverse cardiovascular(CV) event according to whetherthey were assigned to placebo orcaptoprl therapy.

No AdverseCV Events CV Events(n-148) (n-161) (n-69) (n-42)

3. Kostuk WJ, Kazamias TM, Gander MP, Simon AL, Ross J Jr. Leftventricular size after acute myocardial infarction: serial changesand their prognostic significance. Circulation. 1973;47:1174-1179.

4. Ahnve S, Gilpin E, Henning H, Curtis G, Collins D, Ross J Jr.Limitations and advantages of the ejection fraction for defininghigh risk after acute myocardial infarction. Am J Cardiol. 1986;58:872-878.

5. White HD, Norris RM, Brown MA, Brandt PWT, Whitlock RML,Wild CJ. Left ventricular end-systolic volume as the major deter-minant of survival after recovery from myocardial infarction. Cir-culation. 1987;76:44-51.

6. Gruppo Italiano per lo studio della streptochinasi nell'infarto mio-cardico (GISSI). Effectiveness of intravenous thrombolytictreatment in acute myocardial infarction. Lancet. 1986;1:397-401.

7. Sheehan FH, Braunwald E, Canner P, Dodge HT, Gore J, VanNatta P, Passamani ER, Williams DO, Zaret B. The effect ofintravenous thrombolytic therapy on left ventricular function: areport on tissue-type plasminogen activator and streptokinasefrom the Thrombolysis in Myocardial Infarction (TIMI) phase Itrial. Circulation. 1987;75:817-829.

8. Serruys PW, Simoons ML, Suryapranata H, Vermeer F, Wijns WJ,van den Brand M, Bar F, Zwaan C, Krauss XH, Remme WJ, ResJ, Verheugt FWA, van Domburg R, Lubsen J, Hugenholtz PG, forthe Working Group on Thrombolytic Therapy in Acute MyocardialInfarction of The Netherlands Interuniversity CardiologyInstitute. Preservation of global and regional left ventricularfunction after early thrombolysis in acute myocardial infarction.JAm Coll Cardiol. 1986;7:729-742.

9. White HD, Norris RM, Brown MA, Takayama M, Maslowski A,Bass NG, Ormiston JA, Whitlock T. Effect of intravenous strep-tokinase on left ventricular function and early survival after acutemyocardial infarction. N Engl J Med. 1987;317:850-855.

10. Morgan CD, Roberts R, Haq A, Baierie RS, Daly PA, Gent M,Armstrong PW. Coronary patency, infarct size and left ventricularfunction after thrombolytic therapy for acute myocardialinfarction: results from the Tissue Plasminogen Activator: Toronto(TPAT) placebo controlled trial. J Am Coll Cardiol. 1991;17:1451-1457.

11. Jaffe AS, Geltman EM, Tiefenbrunn AJ, Ambos HD, Strauss H,Sobel BE, Roberts R. Reduction of infarct size in patients withinferior infarction with intravenous glyceril trinitrate: a ran-domized trial. Br Heart J. 1983;106:452-460.

12. Pfeffer MA, Pfeffer JM, Steinberg C, Finn P. Survival after anexperimental myocardial infarction: beneficial effects of long-termtherapy with captopril. Circulation. 1985;72:406-412.

13. Jugdutt BI, Warnica JW. Intravenous nitroglycerin therapy to limitmyocardial infarct size, expansion, and complications: effect oftiming, dosage, and infarct location. Circulation. 1988;78:906-919.

14. Pfeffer MA, Lamas GA, Vaughan DE, Parisi AF, Braunwald E.Effect of captopril on progressive left ventricular dilatation afteranterior myocardial infarction. N Engl J Med. 1988;319:80-86.

15. Sharpe N, Smith H, Murphy J, Greaves S, Hart H, Gamble G.Early prevention of left ventricular dysfunction after myocardialinfarction with angiotensin-converting-enzyme inhibition. Lancet.1991;337:872-876.




16. Bonaduce D, Petretta M, Arrichiello P, Conforti G, MontemurroMV, Attisano T, Bianchi V, Morgano G. Effects of captopriltreatment on left ventricular remodeling and function afteranterior myocardial infarction: comparison with digitalis. JAm CollCardiol. 1992;19:858-867.

17. Jugdutt BI, Schwarz-Michorowski BL, Khan MI. Effect oflong-term captopril therapy on left ventricular remodelling andfunction during healing of canine myocardial infarction. JAm CollCardiol. 1992;17:713-721.

18. Jeremy RW, Hackworthy RA, Bautovich G, Hutton BF, Harris PJ.Infarct artery perfusion and changes on left ventricular volume inthe month after acute myocardial infarction. J Am Coll Cardiol.1987;9:989-995.

19. Pfeffer MA, Braunwald E, Moy6 LA, Basta L, Brown EJ, Cuddy TE,Davis BR, Geltman EM, Goldman S, Flaker GC, Klein M, LamasGA, Packer M, Rouleau J, Rouleau JL, Rutherford JR, WertheimerJH, Hawkins CM. Effect of captopril on mortality and morbidity inpatients with left ventricular dysfunction after myocardial infarction:results of the survival and ventricular enlargement trial. N Engi JMedL 1992;327:669-677.

20. Moy6 LA. Central laboratory sampling plans and quality control inclinical trials. Controlled Clin Trials. 1991;12:761-767.

21. Pfeffer MA, Pfeffer JM. Ventricular enlargement and reducedsurvival after myocardial infarction. Circulation. 1987;75(suppl 4):IV-93-IV-97.

22. Erlebacher JA, Weiss JL, Weisfeldt ML, Bulkley BH. Earlydilation of the infarcted segment in acute transmural myocardialinfarction: role of infarct expansion in acute left ventricularenlargement. JAm Coll Cardiol. 1984;4:201-208.

23. Kan G, Visser CA, Koolen JJ, Dunning AJ. Short and long termpredictive value of admission wall motion score in acute myo-cardial infarction: a cross sectional echocardiographic study of 345patients. Br Heart J. 1986;56:422-427.

24. Cleempoel H, Vainsel H, Dramaix M, Lenaers A, Contu E, Hoy-laerts M, Demaret B, de Marneffe M, Vandenbossche J-L, RenardM, Haardt M, Denolin H, Bernard R. Limitations on the prog-nostic value of pre-discharge data after myocardial infarction. BrHeart J. 1988;60:98-103.

25. Kloner RA, Parisi AF. Acute myocardial infarction: diagnostic andprognostic applications of two-dimensional echocardiography. Cir-culation. 1987;75:521-524.

26. Abernethy M, Sharpe N, Smith H, Gamble G. Echocardiographicprediction of left ventricular volume after myocardial infarction.JAm Coll Cardiol. 1991;17:1527-1532.

27. Picard MH, Wilkins GT, Ray PA, Weyman AE. Progressivechanges in ventricular structure and function during the year afteracute myocardial infarction. Am Heart J. 1992;124:24-31.

28. Nishimura RA, Tajik AJ, Shub C, Miller FA, Ilstrup DM, HarrisonCE. Role of two-dimensional echocardiography in the predictionof in-hospital complications after acute myocardial infarction.JAmColl Cardiol. 1984;4:1080-1087.

29. Berning J, Steensgaard-Hansen F. Early estimation of risk by echo-cardiographic determination of wall motion index in an unselectedpopulation with acute myocardial infarction. Am J Cardiol. 1990;65:567-576.

30. Olivetti G, Capasso JM, Sonnenblick EH, Anversa P. Side-to-sideslippage of myocytes participates in ventricular wall remodellingacutely after myocardial infarction in rats. Circ Res. 1990;67:23-34.

31. Pfeffer MA, Braunwald E. Ventricular remodeling after myo-cardial infarction: experimental observations and clinical impli-cations. Circulation. 1990;81:1161-1172.

32. Marino P, Zanolla L, Zardini P. Effect of streptokinase on leftventricular remodeling and function after myocardial infarction:the GISSI trial. JAm Coll Cardiol. 1989;14:1149-1158.

33. Oldroyd KG, Pye MP, Ray SG, Christie J, Ford I, Cobbe SM,Dargie HJ. Effects of early captopril administration on infarctexpansion, left ventricular remodelling and exercise capacity afteracute myocardial infarction. Am J Cardiol. 1991;68:713-718.



Quantitative two-dimensional echocardiographic measurements are major predictors of adverse cardiovascular events after acute myocardial infarction. The protective effects of captopril

Documents