1084

Unstable angina (UA) is the common presenta-tion of a wide variety of underlying diseases, whichmay range from the normal coronary artery to thecomplicated unstable plaque with plaque disrup-tion and thrombus formation.1 Consequently, prog-nosis of patients with this syndrome is very vari-able. Despite the controversy regarding the bestclinical and practical approach to patients withUA, no study to date has definitely demonstratedthat an early invasive management in this settingis always superior to an initially conservative

approach. Therefore risk stratification is recom-mended in these patients to identify those at high-er risk of further events who must undergo a moreaggressive treatment.2,3

Exercise electrocardiography is one of the mostcommonly used tests for risk stratification, and it hasproven useful for this purpose after an acute episodeof UA.4-6 However, recent data have demonstratedthe superiority of exercise echocardiography overexercise electrocardiography in this setting.7

Dobutamine stress echocardiography (DSE) has alsoshown good prognostic value in patients with med-ically stabilized UA.8

Nevertheless, it remains to be seen whether DSEprovides more prognostic information than simpleexercise electrocardiography does in patients withUA who respond to medical therapy. Therefore weperformed a prospective study to compare the prog-nostic value of both DSE and exercise electrocardio-graphy for risk stratification early after an acuteepisode of UA in medically treated patients.

Dobutamine Stress Echocardiography andExercise Electrocardiography for Risk

Stratification in Medically Treated Unstable Angina

Marta Sitges, MD, Manel Azqueta, MD, Carles Paré, PhD, Jordi Magriñá, PhD,Faustino Miranda-Guardiola, MD, Marga Velamazán, RN, Xavier Bosch, PhD, and

Ginés Sanz, PhD, Barcelona, Spain

From the Cardiovascular Institute, Hospital Clínic, Institutd’Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS),University of Barcelona.This work was presented in part at the 10th Scientific Sessions of theAmerican Society of Echocardiography, June 1999, Washington DC.Reprint requests: Carles Paré, Villarroel 170, 08036 Barcelona,Spain (E-mail: cpare@clinic.ub.es).Copyright © 2000 by the American Society of Echocardiography.0894-7317/2000/$12.00 + 0 27/1/107154doi:10.1067/mje.2000.107154

Previous reports have demonstrated the superiority ofexercise echocardiography over exercise electro-cardiography (ex-ECG) for risk stratification inpatients with medically stabilized unstable angina(UA). We sought to analyze the prognostic value ofdobutamine stress echocardiography (DSE) comparedwith ex-ECG for risk stratification in patients with UA.Methods: Ninety-two patients with medically treated UAwere studied (mean age 65 ± 11 years, 24 women, 42%of patients had electrocardiographic abnormalities onadmission). Dobutamine stress echocardiography andtreadmill ex-ECG were performed on the third dayafter hospital admission. End points were recurrentUA, myocardial infarction (MI), or cardiac death.Results: Mean follow-up was 24 ± 7 months. Duringfollow-up, 22 patients had cardiac events (18 recurrentUA, 2 MI, 2 cardiac deaths). The event-free survival rate

was 80% for patients with negative DSE results forischemia and 52% for those with positive DSE results(log rank 9.57; P = .002), compared with an event-freesurvival rate of 79% for patients with negative ex-ECGresults and 66% for those with positive ex-ECG results(log rank 2.06; P = not significant). Left ventriculardysfunction (P = .01) and a positive dobutamine stressechocardiogram (P = .03), but not a positive exerciseelectrocardiogram, were independent predictors ofcardiac events during follow-up.Conclusions: Dobutamine stress echocardiographyperformed early in medically treated patients withUA predicts cardiac events during follow-up moreaccurately and with more specificity than ex-ECGdoes in this population. (J Am Soc Echocardiogr2000;13:1084-90.)

Journal of the American Society of EchocardiographyVolume 13 Number 12 Sitges et al 1085

METHODS

Patients

Patients included in the study were those admitted to ourhospital because of UA,defined as typical chest pain at rest,new onset (less than 1 month) exertional angina of at leastCanadian Cardiovascular Society (CCS) classification classIII in severity,or recent progressive angina (Braunwald classB angina).2 In addition, the last episode of chest pain musthave occurred within 12 hours before admission. Eligiblepatients were those who remained asymptomatic undermedical treatment following the first 48 hours after hospi-tal admission. Exclusion criteria were an increase in crea-tine kinase (CK) and CK-MB plasma levels twice the normalupper limit, evidence of pulmonary edema on admission,recurrent angina during the first 48 hours after admissionor persistent signs of severe ischemia in the electrocardio-gram (symmetrical negative T waves >0.3 mV in 3 or morecontiguous leads), post–myocardial infarction angina orBraunwald class C UA (myocardial infarction [MI] withinthe preceding 2 weeks), secondary angina or Braunwaldclass A UA, and finally, the presence of concomitant severecardiac or noncardiac disease (life expectancy less than 1year). Patients who underwent coronary revascularizationprocedures during the initial hospitalization were alsoexcluded. All patients were treated with aspirin, heparin,and antianginal drugs if no contraindication existed.

An echocardiogram to assess left ventricular functionwas also performed. Patients underwent coronary arteri-ography if clinically indicated. Clinical decisions weremade by the attending physician who considered theresults of the basal echocardiogram and the treadmillstress test. Dobutamine stress echocardiography results forischemia were not available to the attending physician andnot revealed until further analysis had been performed.

Study Protocol

Patients who fulfilled the inclusion criteria underwent DSEand treadmill stress testing within 48 hours after admis-sion. Antianginal drugs, including beta-blocker and calciumchannel blocking agents as well as heparin and aspirin,were not stopped before the tests.

Patients performed a treadmill stress test in accor-dance with the protocol of Bruce et al.9 Heart rate, bloodpressure, and 12-lead electrocardiograms were continu-ously recorded at baseline, during the procedure, andinto recovery. ST-segment deviation was measured 80 msafter the J point. The maximal metabolic equivalents(METS) reached were also determined as an index ofmaximum exercise capacity.A positive result of the testwas considered when angina or downsloping ST-seg-ment depression ≥1 mm in ≥2 contiguous leads devel-oped during the procedure.

Dobutamine stress echocardiography was performedaccording to a modification of the protocol by McNeill etal,10 as described elsewhere.8 A basal echocardiogram wasobtained, and dobutamine was progressively infused.When the target heart rate was not reached, no wallmotion abnormalities developed, and in the absence ofcontraindications, atropine was injected. Causes of testinterruption were the development or worsening of wallmotion abnormalities, 85% of maximum age-predictedheart rate, blood pressure >220/110 mm Hg, severe chestpain, life-threatening arrhythmia, hypotension (systolicblood pressure <90 mm Hg), severe ischemia on the elec-trocardiogram, or intolerable side effects. Interpretationwas made by consensus of 2 experienced echocardiogra-phers unaware of the result of the stress test.The test wasconsidered positive for ischemia when at least one of thefollowing conditions occurred: (1) a new wall motionabnormality appeared in a region with normal rest func-tion, (2) worsening of a rest dyssynergy, or (3) a biphasicresponse of a rest dyssynergy (initial improvement fol-lowed by a worsening in segmental wall motion).

Follow-up

Follow-up data were obtained from telephone interviewsand a review of the patient’s hospital records. End pointswere the combined incidences of cardiac death, nonfatalMI, or recurrent UA after hospital discharge.

Cardiac death in hospitalized patients was determinedthrough documentation of significant arrythmias, congestiveheart failure, or cardiac arrest. Cardiac death was also con-sidered when an unexpected sudden death occurred out ofthe hospital. Nonfatal MI was defined as the presence ofsymptoms and signs of myocardial ischemia with an increasein plasma cardiac enzyme levels twice the normal upperlimit or the development of new Q waves on the rest elec-trocardiogram. Recurrent UA was defined as a new episodeof ischemic symptoms that required a new hospitalization ormedical evaluation (CCS class III-IV angina), without eleva-tion in plasma cardiac enzymes levels, new Q waves in theelectrocardiogram, or new wall motion abnormalities.

Statistical Analysis

Discrete baseline characteristics, clinical outcomes, sensi-tivity, specificity,and accuracy are presented as percentagesand analyzed with the chi-square test. Continuous variablesare expressed as mean value ± SD and compared by theunpaired 2-tailed t test.Multivariate analysis was performedwith a stepwise logistic regression analysis and accordingto a forward selection.Moreover,a multivariate analysis wasalso performed according to a clinically oriented, interac-tive stepwise analysis as previously reported.11 In this inter-active analysis, variables were included in the model in thesame order in which they are actually considered by an

Journal of the American Society of Echocardiography1086 Sitges et al December 2000

attending physician. Event-free survival was described withthe Kaplan-Meier product limit method, with differencesbetween groups compared with the log-rank chi-square sta-tistic. A P value <.05 was considered statistically significant.

RESULTS

Patients

From August 1996 to March 1997, a total of 290patients were admitted because of UA. Of them, 132patients fulfilled the study’s inclusion criteria. Ten(7%) of the 132 patients who met the inclusion cri-teria could not be studied with DSE because of pooracoustic image (8 patients, 6%), apical thrombus (1patient, 0.7%), or a history of sustained ventriculartachycardia (1 patient, 0.7%). Dobutamine stressechocardiography was performed in the remaining122 patients (test feasibility 94%).On the other hand,only 92 of the initially included 132 patients wereable to perform an exercise test (feasibility 69%).

Mean age was 65 ± 11 years (range 36 to 89 years)and 24 subjects were women (26%). Twenty-sixpatients (28%) had a history of MI. Twelve patientswere admitted because of progressive exertional angi-na (13%) and the other 80 (87%) had angina at rest.Fifty-eight patients (63%) had angina that lasted >20minutes,and 77 (97%) came to the hospital within thefirst 48 hours after the onset of symptoms.About 50%of the patients had electrocardiographic (ECG) abnor-malities on admission: 60% of them had ST-segmentdepression, and 40% had T-wave inversion. Seventy-two patients (78%) received beta-blocker therapy; theother patients received non-dihydropyridinic calcium

channel blockers. All the patients were treated withaspirin and heparin if no contraindication existed;these drugs continued to be administered during DSEperformance (Table 1).

Tests Results

No major complications (death, MI, or life-threaten-ing arrhythmia) occurred during DSE or exercise testperformance. No differences were found in maxi-mum blood pressure nor in heart rate achieved dur-ing DSE and exercise electrocardiography (162 ± 30versus 154 ± 24 mm Hg,P = not significant [NS]; and109 ± 24 versus 117 ± 19, P = NS; respectively).Target heart rate, defined as 85% of theoretic maxi-mum heart rate (220 minus age in years), was onlyreached in 19% of the patients in the treadmill stresstest as compared with 25% in the DSE test (P = NS).Mean maximum workload and mean exercise dura-tion at the treadmill exercise test were 7.8 ± 2.3METS and 6 ± 2 minutes, respectively.

Positive DSE results for ischemia were obtained in20 patients (22%); the other 72 patients (78%) hadnegative DSE results for ischemia. Only 12 patients(10%) developed ECG changes during the test, all ofwhich consisted of ST depression. None of thepatients showed ST elevation. Regarding exerciseelectrocardiography, 34 patients (37%) had positiveresults for ischemia, whereas the other 58 patients(63%) had negative results.

Follow-up Data

Follow-up was obtained in all patients (mean 24 ± 7months). After this time, 22 patients (24%) sufferedcardiac events: 2 of whom (2%) died, 2 (2%) had anMI, and 18 (10%) were readmitted because of UA. Ofthe 20 patients with positive DSE results, 9 (45%) hadclinical events:2 patients died,1 had an MI,and 6 werereadmitted with UA. On the other hand, 13 (18%) of72 patients with negative DSE results had clinicalevents: 12 had recurrent UA, and 1 had an MI; therewere no cardiac deaths (specificity 85%, negative pre-dictive value 81%). Regarding exercise electrocardiog-raphy, of the 34 patients with a positive test, 11 (32%)had cardiac events during follow-up: 1 patient died, 1patient had an MI, and 9 were readmitted because ofUA.Of the 58 patients with a negative exercise test,11(19%) had cardiac events during follow-up: 1 patientdied,1 patient had an MI,and 9 patients had recurrentUA (specificity 69%, negative predictive value 81%).

The event-free survival rate of patients with nega-tive DSE results was 80% compared with 52% forthose with positive DSE results (log rank 9.57, P =.002).However,exercise electrocardiography did notdifferentiate the two groups of patients: survival

Table 1 Baseline characteristics (n = 92)

Age 65 ± 11 yearsWomen 24 (26%)Hypertension 46 (50%)Diabetes mellitus 18 (19.5%)Prior myocardial infarction 26 (28%)Progressive exertional angina 12 (13%)Rest angina: 80 (87%)

<48 hours from onset 77 (97%)>48 hours from onset 3 (3%)>20 minutes duration 58 (63%)

ECG changes: 39 (42%)ST depression 23 (25%)Negative T wave 16 (17%)

Beta-blocker therapy 72 (78%)Aspirin 91 (99%)Heparin 84 (91%)

Data presented are mean ± SD or number (%) of patients. ECG changes: STdepression or negative T wave on rest electrocardiogram or during chest pain.

Journal of the American Society of EchocardiographyVolume 13 Number 12 Sitges et al 1087

without events was 79% for patients with a negativeexercise electrocardiogram and 66% for those with apositive test (log rank 2.06, P = NS) (Figure 1).Moreover, the survival rate for patients without an MIwas 98% for patients with negative DSE results com-

pared with 83% for those with positive DSE results(log rank 9.03, P = .002), whereas it was 96% and94% for patients with negative and positive exerciseelectrocardiograms, respectively (log rank 0.34, P =NS) (Figure 2).

Figure 1 Event-free survival rates of patients with positive or negative dobutamine stress echocardiogra-phy (DSE) and exercise electrocardiography (Ex-ECG) results for ischemia. DSE–, Negative DSE results;DSE+, positive DSE results; Ex-ECG–, negative Ex-ECG results; Ex-ECG+, positive Ex-ECG results; LR,log rank; NS, not significant.

Figure 2 Survival without myocardial infarction of patients with positive or negative dobutamine stressechocardiography (DSE) and exercise electrocardiography (Ex-ECG) results for ischemia. DSE–,Negative DSE results; DSE+, positive DSE results; Ex-ECG–, negative Ex-ECG results; Ex-ECG+, posi-tive Ex-ECG results; LR, log rank; NS, not significant.

Journal of the American Society of Echocardiography1088 Sitges et al December 2000

Univariate and Multivariate Analysis

Univariate predictors of cardiac events during fol-low-up were a history of prior MI, ECG changes onadmission, left ventricular ejection fraction, and pos-itive DSE results (Table 2); exercise electrocardiog-raphy test results were not predictive of further car-diac events in this group of patients. Multivariateanalysis selected as independent predictors of car-diac events left ventricular dysfunction (left ventri-cle ejection fraction <45%) and positive DSE resultsfor ischemia (Table 3). When individual variableswere entered in the logistic regression modelaccording to an interactive clinically orientedapproach, ECG changes on admission, left ventricu-lar dysfunction, and positive results of DSE were

independent and additional predictors of cardiacevents during follow-up (Table 4).

DISCUSSION

In this study, DSE performed early after an acuteepisode of UA in medically treated patients was abetter and more specific predictor of cardiac eventsduring follow-up than exercise electrocardiography,especially when considering the incidence of MI andcardiac death.

The prognosis of patients who have an acuteepisode of UA is very variable, with incidences of MIand death at 1 year after hospital admission rangingfrom 13% to 25%.12,13 Although the optimal manage-ment approach for these patients is still open todebate,14,15 risk stratification is usually recommend-ed to identify those patients at a higher risk of fur-ther cardiac events who must undergo a moreaggressive treatment.2,3 Previous studies have shownthat exercise electrocardiography is safe and usefulfor risk stratification in patients with UA.4-6,16 It isalso known that patients with UA who initiallyrespond to medical therapy without recurrence ofsymptoms have a better prognosis.17 However, it hasbeen reported recently that up to 20% of the patientswith UA who are medically treated and have negativeexercise test results die or suffer an MI during the fol-lowing 5 years.18Therefore data of exercise test para-meters may not be enough to establish a prognosisin this group of patients. Clinical variables such asmale sex, the presence of diabetes mellitus, and priorMI have been proven to be independent predictorsof cardiac events during follow-up in this setting.18

In our study, history of a previous infarction was apredictor of subsequent events but failed to reachstatistical significance on the multivariate analysis,probably because of the small sample size.Nevertheless, left ventricle dysfunction and positiveDSE results for ischemia were selected as indepen-dent predictors of cardiac events during the 2 yearsafter hospital admission. Also in accordance withother studies,7,18 the results of the exercise electro-cardiography test did not identify those patients whowould suffer events during follow-up.

Both exercise echocardiography and DSE havedemonstrated their safety and usefulness for riskstratification early after an acute episode of UA.7,8

Early data suggested that exercise echocardiographywas useful and even better than exercise electrocar-diography for that purpose.19,20 These initial datawere confirmed in a more recent and larger study,showing that exercise echocardiography, but not

Table 2 Characteristics of patients according to clinicaloutcomes: univariate analysis

Cardiac Noncardiacevents events P

(n = 22) (n = 70) value

Age (y) 65 ± 12 65 ± 11 NSWomen 3 (13%) 19 (27%) NSDiabetes mellitus 4 (18%) 18 (25%) NSHypertension 8 (36%) 14 (20%) NSPrior MI 11 (50%) 15 (21%) .001ECG changes 14 (63%) 8 (11%) .02

on admissionLVEF (%) 49 ± 11 56 ± 10 .02Positive DSE 9 (41%) 11 (15%) .01Positive exercise 11(50%) 23 (33%) NS

electrocardiogram

Data presented are mean ± SD or number (%) of patients. NS, Not signifi-cant; MI, myocardial infarction; ECG, electrocardiographic; LVEF, left ven-tricular ejection fraction; DSE, dobutamine stress echocardiography.

Table 3 Multivariate analysis after a forward procedure

Odds ratio 95% Confidence limits P value

LVEF <45% 4.1 1.3 - 12 .01Positive DSE 3.7 1.2 - 10.7 .03

LVEF, Left ventricular ejection fraction; DSE, dobutamine stressechocardiography.

Table 4 Multivariate analysis after an interactive procedure

Odds ratio 95% Confidence limits P value

Prior MI 1.4 0.35 - 5.5 .6ECG changes 3.2 1.09 - 9.4 .03LVEF <45% 3.6 1.2 - 11.2 .02Positive DSE 3.5 1.1 - 11.4 .03

MI, Myocardial infarction; ECG changes, electrocardiographic changes onadmission (ST depression or T wave inversion); LVEF, left ventricular ejec-tion fraction; DSE, dobutamine stress echocardiography.

Journal of the American Society of EchocardiographyVolume 13 Number 12 Sitges et al 1089

exercise electrocardiography, was predictive of sub-sequent events in medically treated patients withUA.7 In addition, we also recently demonstrated thatDSE is feasible and safe when performed early afteran acute episode of UA in a low- to intermediate-riskpopulation. In this group of patients, DSE proved tobe a good predictor of further cardiac events.8

Few data exist comparing the role of exercise elec-trocardiography and DSE in patients with UA. Dipy-ridamole stress echocardiography proved to be ofincremental prognostic value compared with exerciseelectrocardiography in a study of 429 patients withsuspected coronary artery disease.11 Actually, bothdipyridamole and dobutamine stress echocardiogra-phy have shown similar prognostic yields in risk strati-fication of low- to moderate-risk patients with knownor suspected coronary artery disease.21 Recently,Desideri et al22 reported similar predictive values ofexercise electrocardiography and pharmacologic(dobutamine or dipyridamole) stress echocardiogra-phy in patients with noncomplicated non-Q wave MI.

The prognostic value of stress echocardiography inpatients with UA may be reduced compared with thatin other clinical settings because of the underlyingpathophysiology of acute coronary syndromes morerelated to plaque rupture than plaque hemodynam-ics.1 However,our results are in accordance with pre-vious exercise echocardiography studies that showedthat echocardiography has a better predictive valuethan exercise electrocardiography. Importantly, DSEbut not exercise electrocardiography identified thosepatients who died or suffered an MI during follow-up.On the other hand, specificity was higher for DSEthan for exercise electrocardiography.

Even though exercise parameters such as maxi-mum workload have been reported to be of prognos-tic usefulness,4-6 the high rate of beta-blocker therapyamong our patients prevented them from reachingtarget heart rate and experiencing increased bloodpressure during the stress test.This applied both forthe treadmill stress test and the DSE test, affectingboth prognostic capabilities in the same manner.

Finally, in our study, DSE was more feasible thanexercise electrocardiography. This is becoming animportant factor for consideration because the pro-portion of persons who cannot exercise (eg, elderlypatients and patients with disabilities) is increasingamong those admitted to the hospital because of UA.Accordingly, DSE appears to be a more appropriatetest in such circumstances, providing good informa-tion about the probability of further cardiac eventsin these patients.

Nevertheless, there are some limitations of thestudy that deserve to be pointed out. First, this study

was performed with a selected low- to intermediate-risk population of patients with UA. Moreover,because of the study design, the attending physicianwas blinded to the results of the DSE test,and the DSEreader was blinded to the results of the treadmillstress test, leading to a small bias in favor of DSE forthe assessment of prognosis. Consequently, theseresults cannot be applied to all patients with UA, butthey may be useful for those with negative exerciseelectrocardiography test results who are still at con-siderable risk of further cardiac events.18 Second,eachmedical center, aware of its own availabilities and lim-itations, must take the best approach to the manage-ment of patients with UA because imaging studiesmay be overused and echocardiography laboratoriesmay be overbooked. Despite this fact, the initial andclinical selection of the patients may result in a rela-tively small number of patients who would not repre-sent a significant volume for an experienced echocar-diography laboratory. Actually, in our study,the eligiblepatients only represented 31% of the total number ofpatients admitted to the hospital because of UA duringthe inclusion period. Finally, the small sample ofpatients studied prevents the establishment of definiteconclusions,but initial data show a role for DSE in fur-ther risk stratification of patients with medically treat-ed UA who have negative exercise stress test results.

REFERENCES

1. Fuster V, Badimon L, Badimon JJ, Chesebro JH. The patho-genesis of coronary artery disease and the acute coronary syn-dromes. N Engl J Med 1992;326:242-50.

2. Braunwald E, Mark DB, Jones RH, et al. Unstable angina:diagnosis and management: clinical practice guideline.Agency for Healthcare Policy and Research and the NationalHeart, Lung and Blood Institute, Public Health Service, USDepartment of Health and Human Services. Rockville (MD):AHCPR Publication 1994, No. 94-0602:154. p. 28 and 92.

3. Theroux P, Fuster V. Acute coronary syndromes: unstableangina and non-Q wave myocardial infarction. Circulation1998;97:1195-206.

4. Swahn E, Areskog M, Berglund U, Walfridsson H, WallentinL. Predictive importance of clinical findings and a predis-charge exercise test in patients with suspected unstable coro-nary artery disease. Am J Cardiol 1987;59:208-14.

5. Wilcox I, Freedman SB, Allman KC, Collins FL, Leitch JW,Kelly DT, et al. Prognostic significance of a predischargeexercise test in risk stratification after unstable angina pec-toris. J Am Coll Cardiol 1991;18:677-83.

6. Lindahl B, Andren B, Ohlsson J, Venge P, Wallentin L, forthe FRISC Study Group. Noninvasive risk stratification inunstable coronary artery disease: exercise test and biochemi-cal markers. Am J Cardiol 1997;80:40E-44E.

7. Lin SS, Lauer MS, Marwick TH. Risk stratification of patientswith medically treated unstable angina using exerciseechocardiography. Am J Cardiol 1998;82:720-4.

Journal of the American Society of Echocardiography1090 Sitges et al December 2000

8. Sitges M, Paré C, Azqueta M, Bosch X, Miranda-GuardiolaF, Velamazán M, et al. Feasibility and prognostic value ofdobutamine-atropine stress echocardiography early in unsta-ble angina. Eur Heart J 2000;21:1063-71.

9. Bruce RA, Kusumi F, Hosmer D. Maximal oxygen intake andnomographic assessment of functional aerobic impairment incardiovascular disease. Am Heart J 1973;85:546-62.

10. McNeill AJ, Fioretti PM, el-Said SM, Salustri A, Forster T,Roelandt JR. Enhanced sensitivity for detection of coronaryartery disease by addition of atropine to dobutamine stressechocardiography. Am J Cardiol 1992;70:41-6.

11. Severi S, Picano E, Michelassi C, Lattanzi F, Landi P,Distante A, et al. Diagnostic and prognostic value of dipyri-damole echocardiography in patients with suspected coronaryartery disease: comparison with exercise electrocardiography.Circulation 1994;89:1160-73.

12. Mulcahy R, Daly L, Graham L, Hickey N, O’Donoghue S,Owens R, et al. Unstable angina: natural history and deter-minants of prognosis. Am J Cardiol 1981;48:525-8.

13. Betriu A, Heras M, Cohen M, Fuster V. Unstable angina:outcome according to clinical presentation. J Am CollCardiol 1992;19:1659-63.

14. TIMI III-B Investigators. Effects of tissue-plasminogen acti-vator and a comparison of early invasive and conservativestrategies in unstable angina and non-Q-wave myocardialinfarction. Circulation 1994;89:1545-56.

15. The FRISC Investigators. Invasive compared with non-inva-

sive treatment in unstable coronary artery disease: FRISC IIprospective randomised multicentre study. Lancet 1999;354:708-15.

16. Buntman SM, Olson HG, Gardin JM, Piters KM, Hullett M,Buntman LK. Submaximal exercise test after stabilization ofunstable angina pectoris. J Am Coll Cardiol 1985;4:667-73.

17. Castañer A, Roig E, Serra A, Flores T, Magriña J, Azqueta M,et al. Risk stratification and prognosis of patients with recentonset angina. Eur Heart J 1990;11:868-75.

18. Moreno R, Lopez de Sá E, López-Sendón JL, Ortega A,Fernández MJ, Fernández-Bobadilla J, et al. Prognosis ofmedically stabilized unstable angina pectoris with a negativeexercise test. Am J Cardiol 1998;82:662-5.

19. Amanullah AM, Lindvall K. Predischarge exercise echocar-diography in patients with unstable angina who respond tomedical therapy. Clin Cardiol 1992;15:417-23.

20. Eriksson SV, Erhardt L, Lindvall K, Melcher A, Rehnquist N.Long-term prognostic importance of exercise echocardiographyafter an episode of unstable angina. Cardiology 1995;86:426-31.

21. Pingitore A, Picano E, Varga A, Gigli G, Cortigiani L,Previtali M, at al. Prognostic value of pharmacological stressechocardiography in patients with known or suspected coro-nary artery disease. J Am Coll Cardiol 1999;34:1769-77.

22. Desideri A, Bigi R, Suzzi GL, Coletta C, Gregori D, ValenteG, et al. Stress echocardiography and exercise electrocardiog-raphy for risk stratification after non-Q-wave uncomplicatedmyocardial infarction. Am J Cardiol 1999;84:739-41.

Receive tables of contents by e-mail

To receive the tables of contents by e-mail, sign up through our Web site at

http://www.mosby.com/echo

Choose E-mail Notification. Simply type your e-mail address in the box and click theSubscribe button.

Alternatively, you may send an e-mail message to majordomo@mosby.com. Leave thesubject line blank, and type the following as the body of your message:

subscribe echo_toc

You will receive an e-mail message to confirm that you have been added to the mailinglist. Note that table of contents e-mail messages will be sent when a new issue is postedto the Web site.