Prognostic significance of precordial ST-segment depression during inferior acute myocardial infarction

Prognostic Significance of Precordial ST-Segment Depression During Inferior Acute Myocardial Infarction MARK A. HLATKY, MD, ROBERT M. CALIFF, MD, KERRY L. LEE, PhD, DAVID B. PRYOR, MD, GALEN S. WAGNER, MD, and ROBERT A. ROSATI, MD

To study the mechanism and prognostic importance of precordial ST-segment depression during inferior acute myocardial infarction, 162 patients admitted during 1969 through 1982 were identified. Patients with ST depression in leads VI, V2 and Vs had significantly larger infarctions as assessed by a QRS scoring system. Hospital mortality was 4 % (3 of 75) among patients without ST depression, and 13 % (11 of 87) in patients with ST depression. The relation between the amount of ST depression and hospital mortality was significant (p
adjusting for other potentially prognostic factors. Among patients discharged from the hospital, the 5-year survival was 92% in those without precordial ST depression and 80 % in those with precordial ST depression (p = 0.058 by the Cox model). Precordial ST-segment depression on the admission electrocardiogram during an inferior acute myocardial infarction indicates a larger infarction, prediets a higher hospital mortality and suggests a worse long-term prognosis after discharge.

The significance of the precordial ST-segment depression that sometimes accompanies inferior acute myocardial infarction (AMI) remains controversial. This reciprocal change may result passively from inferior ST-segment elevation,r but several studies suggest that it may represent an active, primary process. In particular, it may indicate posterior epicardial injury and a larger area of jeopardized myocardium, consistent with the higher peak levels of creatine kinase and lower ejection fractions found in such patients.2-6 Alternatively, this electrocardiographic (ECG) change may arise from anterior wall ischemia,7 since some studies have found an association with left anterior descending stenosis,8Tg although others have not.4s10Thus, precordial ST depression during inferior AM1 may be a benign

ECG finding of little clinical importance, or it may be a valuable marker of patients with increased risk. The major goal of this study was to define the prognostic significance of the precordial ST-segment depression in the admission electrocardiogram of patients with inferior AMI. Since many other factors also affect prognosis after AMI, multivariable analyses were used to determine if precordial ST changes contribute independent information. The mechanism of precordial ST-segment changes was also investigated by identifying the location and estimating the extent of the acute infarct, as indicated by the QRS changes.

(Am J Cardiol 1985;55:325-329)

Methods Since January 1969, data from all patients admitted to the Duke University cardiac care unit with AM1 have been prospectively entered into a computer database, described in detail elsewhere.ll Information on 52 clinical variables is recorded daily and stored. Patients have been followed after discharge at 6 months, at 1 year and annually thereafter. All patients who met the following criteria were identified: a history of prolonged chest pain, characteristic changes in serum enzymes, and development of a Q wave 0.03 second or longer in lead aVF. Patients admitted to the cardiac care unit by interward or interhospital transfer were excluded. Patients with either history or ECG evidence of previous MI or ECG criteria for ventricular hypertrophy or bundle branch block were also excluded, as were those with ST elevation in leads VI, Vz or Vs. Patients with some ST-segment elevation in leads Vq, V5 and Vs associated with ST depression in VI, Vz

From the Division of Cardiology, Department of Medicine, and the Division of Biometry, Department of Community and Family Medicine, Duke University Medical Center, Durham, North Carolina. This work was supported by Research Grant HS 04873 from the National Center for Health Services Research, Rockville, Maryland; Research Grant HL 17670 from the National Heart, Lung, and Blood Institute, Bethesda, Maryland; Training Grant LM 07003 and Grants LM 03373 and LM 00042 from the National Library of Medicine, Bethesda, Maryland; a grant from the Prudential Insurance Company of America, Newark, New Jersey; a grant from the Kaiser Family Foundation, Menlo Park, California; and a grant from the Andrew W. Mellon Foundation, New York, New Ycrk. Manuscript received May 14, 1984; revised manuscript received November 5, 1984, accepted November 6, 1984. Address for reprints: Mark A. Hlatky, MD, P.O. Box 31265, Duke University Medical Center, Durham, North Carolina 27710. 325

326

RECIPROCAL

CHANGE

AND INFERIOR

INFARCTION

PROGNOSIS

and Vs were accepted, however, if the amount of this ST elevation was less than that in inferior leads II, III and aVF, since this was considered indicative of inferoapical involvement.12,13 Patients seen after June 1982were excluded to ensure at least 1 year of follow-up for each patient. Admission electrocardiograms were reviewed by cardiologists blinded to the follow-up information. ST-segment deviation was identified by a non-isoelectric J point. The amount and direction of any ST-segment deviation from the isoelectric baseline was measured at 0.08 second after the J point to the nearest 0.05 mV in all 12 leads. For purposes of analysis, the amount of ST-segment depression in leads Vi, Vs and Vs was summed. Prognosis: The effect of precordial ST depression in Vi, Vs and Vs was examined for both short-term prognosis (hospital mortality) and long-term prognosis. To use all the information available, ST depression was used mstatistical tests as a continuous variable rather than a categorical “present” or “absent.” l4 For easein describing the findings, however, the patients were divided into 2 groups. Patients in group 1 had less than 0.05 mV of ST depression in each of leads Vi, Vz and Vs, and the remaining patients formed group 2. The logistic regression model was used to test whether the sum of precordial ST-segment depression in leads VI, Vs and Vs was predictive of hospital survival. Breslow’s formulation15 of the Cox proportional hazards model16 was used to test whether the sum of precordial ST depression was predictive of long-term prognosis in hospital survivors. Initially, univariable analyses were performed. Subsequently, multivariable analyses were performed to test whether the sum of precordial ST-segment depression added significantly to the prognostic information provided by other factors. The logistic regression model was also used to test whether the sum of precordial ST-segment depression in leads Vi, Vs and Vs was predictive of urgent complications during hospitalization. The occurrence of 1 or more of the following was considered an urgent complicationr7: ventricular tachycardia, ventricular fibrillation, persistent hypotension, Killip class III or IV, or reinfarction. Ventricular tachycardia was defined as 3 or more beats of ventricular origin at a rate of 100 beats/ min or greater. Persistent hypotension was defined as a systolic,blood pressure less than 90 mm Hg lasting for 1 hour or more. Comparison of QRS changes: The pathogenesis of precordial ST depression was investigated by using the predischarge electrocardiogram to judge the size and extent of infarction by means of the QRS scoring system.ls,is This scoring system consists of criteria regarding Q- or R-wave duration and amplitude ratios for each of 10 leads (I, II, aVL, aVF and Vi through Vs). The accuracy of this method in predicting the size and location of myocardial necrosis has been validated.12,1s We hypothesized that (1) if precordial ST depression were the result of posterior injury, patients with this finding should show QRS changes of posterior infarction more frequently; and (2) if precordial ST depression were due to anterior subendocardial injury, patients with this finding should develop QRS changes of anterior infarction more frequently. To test the first hypothesis, the number of QRS points in leads V1 and Vz indicative of posterior wall damage developed by patients without precordial ST depression (group 1) was compared with the number of posterior wall QRS points developed by patients with precordial ST depression (group 2). Similarly, to test the second hypothesis, the number of anterior wall QRS points in leads Vi through V4 (development of Q waves or R-wave loss) developed by group 1 patients was compared to those developed by group 2 patients. Statistical significance

was assessedby means of nonparametric l-way analysis of variance (Kruskal-Wallis test). Results

Prognosis: A total of 162 consecutive patients (120 men, 42 women) met entry criteria for the study. The median age was 58 years (range 33 to 89). Fourteen patients (8.6%) died in hospital and 32 more died from cardiovascular causesin the long-term follow-up period (median 4 years, range 1 to 10). Hospital mortality was 4% (3 of 75) among patients with less than 0.05 mV of ST-segment depression in leads Vi, Vs and Vs (group l), compared with 13% (11 of 87) in patients with at least 0.05 mV of ST depression in these leads (group 2). The sum of ST-segment depression was significantly related to hospital.mortality when tested with the logistic model (p
February 1, 1985

The analysis of complications was repeated after excluding the patients who died during the hospitalization. Urgent, nonfatal complications occurred in 21 of the 72 patients (29%) without precordial ST depression and in 35 of the 76 patients (46%) with precordial ST-segment depression. The sum of precordial ST depression was significantly related to the probability of developing urgent, nonfatal complications (p = 0.026 by the logistic model), This relation was of borderline statistical significance (p = 0.055) after adjustment for other prognostic factors, probably becauseof the smaller sample size after exclusion of hospital deaths. Survival after hospital discharge also differed depending on the sum of precordial ST-segment depression present on the admission ECG (Fig. 1). The survival at 1 year, 3 years and 5 years, calculated by the Kaplan-Meier method, was 96%, 94% and 92%,respectively, in group 1, and 89%, 86% and 80%, respectively, in group 2. The sum of precordial ST-segment change was a prognostic factor of borderline statistical significance, when tested univariately with the Cox model (p = 0.058). Multivariable analyses of long-term prognosis were not performed because of the borderline significance in the univariate analyses. Comparison of QRS changes: The patients in groups 1 and 2 were compared to examine whether precordial ST-segment change was associated with a different extent or location of infarction by means of the QRS scoring system. 12,13QRS scores could not be calculated in 3 of the 75 patients from group 1 and 4 of the 87 patients from group 2 because of development of bundle branch block, insertion of a pacemaker or lack of follow-up electrocardiograms. The median interval between the admission and predischarge electrocardiograms was 11 days. Groups 1 and 2 had distinctly different patterns of infarction as judged by the QRS scoring system (Table I). Patients without ST depression (group 1) had significantly fewer overall QRS points (p = 0.041 by Kruskal-Wallis test), suggesting that the total area of infarction was smaller. Groups 1 and 2 were no different, however, in the number of QRS points developed in leads reflecting the inferior wall. If precordial ST depression were the result of anterior ischemia, a higher frequency of associated anterior infarction would be expected. There was, however, no significant difference in the amount of anterior QRS change between the 2 groups. Supporting the hypothesis that precordial ST depression indicates posterior epicardial injury was the observation that patients in group 2 had significantly greater amounts of posterior infarction (p = 0.0012). The amount of apical infarction was also significantly increased (p = 0.0086) when ST elevation in Vq, Vs and Vs accompanied ST depression in leads Vi, Vs and Va. These differences in the distribution of QRS points was not changed appreciably by excluding the 10 patients who had reinfarction before hospital discharge. Discussion The major finding of this study was that patients with precordial ST-segment depression during an inferior

THE AMERICAN

TABLE I

JOURNAL

OF CARDIOLOGY

Volume 55

327

Location and Mean QRS Points Relative to Precordial ST-Segment Depression Precordial ST Depression

Leads inferior (II, aVF) Posterior (V,,V,) Anterior (VI-V,) Apical (Vs-Vs) Lateral (I, aVL) Total

Absent (Group 1, n = 72)

Present (Group 2, n = 83)

p Value

3.6 0.5 0.3 0.4 0.0 4.8

3.6 1.0 0.4 0.8 0.1 5.9

0.880 0.001 0.701 0.009 0.506 0.041

AM1 have a poorer prognosis. Patients with precordial ST depression had a hospital mortality rate of 13%, compared with 4% in the remaining patients. The risk of hospital death increased directly with the sum of ST depression, without evidence for a threshold, and the prognostic information it provided was independent of that provided by clinical factors and by the amount of inferior ST elevation. Finally, precordial ST depression was also predictive of urgent complications during the hospitalization among patients who survived. Previous studies of precordial ST depression as a predictor of hospital mortality have yielded conflicting results. In the initial study by Shah et a1,2hospital mortality was 21% (5 of 24) in patients with precordial ST depression, compared with 0% (0 of 20) in patients without ST depression. No other study has found as high a mortality rate in patients with precordial ST depression, and none had confirmed a significant difference in hospital mortality.s-@Js The present investigation differs from previous studies in 2 important respects. First, most reported studies have included fewer than 50 patients, and none has included more than 110 patients with inferior AMI.

0 = NO ST DEPRESSION : ST DEPRESSION

l

t 2058 52 4

49 47 1

33 39 1

30

1

67 57 I

1

2

3 YEARS

4

5

6

ON:71

l N:67

32 I

FIGURE 1. Cumulative survival of patients discharged from the hospital, according to the presence or absence of precordial ST-segment depression on the admission electrocardiogram.

328

RECIPROCAL

CHANGE

AND INFERIOR INFARCTION

PROGNOSIS

The small sample sizes may have prevented previous studies from detecting real mortality differences. In most previous studies an insignificant trend toward higher hospital mortality related to precordial ST depression was present.2~~~8~g*20 Second, the present study analyzed precordial ST depression as a continuous variable, rather than as simply “present” or “absent,” to use all available information and increase statistical power.14Smaller sample sizes and insensitive statistical methods are particularly likely to mask clinically important differences in mortality. Our results also suggest that the long-term prognosis of patrer :.::with precordial ST depression is worse than that of p&ients without this finding (Fig. 1). Among patients discharged from the hospital, the &year survival was 80% in patients with and 92% in patients without ST depression (p = 0.058 by Cox analysis of the entire survival curve). Previous studies have reported follow-up data after discharge from the hospital, but none has documented important differences in longterm mortality after discharge.P6J8JgThe largest study, that of Wasserman et al,19found equal mortality 3 years after inferior myocardial infarction whether or not precordial ST depression was present among 219 patients enrolled in the Aspirin Myocardial Infarction Study. However, patients in that study were highly selected and may not be representative of all patients with inferior infarction; all were Killip class I or II during hospitalization, and all had survived for at least 2 months after discharge before enrollment. The present study differs from previous studies by having more patients, or longer follow-up, or both. Pathogenesis of ST depression: The mechanism of precordial ST depression remains controversial. ST depression might occur as a passive “reciprocal” phenomenon. We found, as others have, that the sum of ST elevation in the inferior leads II, III and aVF was correlated with the sum of precordial ST depression in leads Vi, Vz and Vs. This correlation need not imply that precordial ST depression is a benign electrical phenomenon, however. Precordial ST depression was a stronger predictor of prognosis than inferior ST elevation, and the latter was not significant in the multivariable analysis. Precordial ST depression may indicate a larger infarction. We examined this possibility by means of the QRS scoring system that correlates well with the location and extent of myocardial infarction.12J3 Patients with precordial ST depression had larger total QRS scores, suggesting that they had larger infarctions (Table I). The size of the inferior infarct itself was similar, however, whether or not precordial ST change was present. Patients with ST depression in leads Vi, Vz and Va had associated posterior wall and apical infarction significantly more often than patients without this ECG finding. In contrast, anterior and lateral infarctions were not associated with precordial ST depression. These observations suggest that precordial ST changes mark a group of patients with larger infarcts that often include the true posterior wall and apex. The present study was not designed primarily to investigate the mechanism of precordial ST depression.

Data from additional ECG leads, coronary angiography and ventriculography, which could have shed further light on the pathophysiologic mechanisms, were not available on most subjects. Our observations concerning the correlations between ST changes and QRS changes on the electrocardiogram are in broad accord with the findings of other investigators using more sophisticated techniques. Most previous investigations have also found evidence of larger myocardial infarctions in patients with precordial ST depression. The peak creatine kinase level is usually higher in patients with precordial ST depression,2-6 although this has not been a uniform finding. I7 Larger areas of inferior infarction in association with precordial ST-segment depression have also been disclosed by thallium-201 perfusion imaging.436 Global left ventricular ejection fraction has usually been found to be lower in patients with precordial ST depression,2-4T6with 1 excepti0n.l Several studies have examined regional wall motion abnormalities and correlated them with precordial ST change. These studies are difficult to compare because of variations in technique and nomenclature. Most found more extensive inferior wall motion abnormalities,2>416but disagree on the presence of other regional abnormalities. Some studies found no difference in wall motion outside the infarct zone,1’4p6 while others found additional anteroseptal,2 apical3 or posterolatera13 abnormalities. Another potential cause of precordial ST depression is anterior ischemia due to concomitant disease of the left anterior descending coronary artery. In our study, evidence of associated anterior infarction was not more common in patients with precordial ST depression. Previous investigations disagree whether an association exists between left anterior descending disease and precordial ST depression, with some studies finding an association,8,g while others have not.4J0 Scant evidence for anterior wall ischemia has been found with either thallium-201 perfusion imaging@ or positron emission tomography. 2o The data are insufficient to establish anterior ischemia as the cause of most cases of precordial ST depression during inferior AMI, but cannot exclude it as a contributing factor in some cases. Implications: Therapeutic decisions in patients during and after AM1 depend on accurate assessment of prognosis, and precordial ST changes during inferior AM1 may provide a valuable noninvasive method of classifying patients into high- and low-risk groups. Information from the admission electrocardiogram is available immediately, at no risk and little cost, in even the smallest community hospital. In high-risk patients, interventions aimed at reducing risk may be appropriate, while in low-risk patients, conservative therapy and rapid rehabilitation may be more suitable. Precordial ST changes mark patients at high risk of hospital complications or death, probably because of a larger area of infarction. Thrombolytic therapy is being investigated as a method of salvaging acutely jeopardized myocardium, and may be particularly indicated in patients with precordial ST depression during inferior AMI. While similar prognostic information may be

February I,1985

available from other methods (such as perfusion scintigraphy or radionuclide angiography), the immediate availability of such information from the admission electrocardiogram is an obvious advantage when rapid decisions must be made. Acknowledgment: We thank Bernie McCants and Kathleen Coffey for technical assistance, Cristy Vollmar for typing the manuscript, and Daniel B. Mark, MD, and Joseph C. Greenfield, Jr., MD, for their critical review of the manuscript.

9.

IO. 11. 12.

References 13. 1. Croft CH, Woodward W, Nicod P, Corbetl JR, Lewis SE, Willerson JT, Rude RE. Clinical implications of anterior S-T segment depression in patients with acute inferior myocardial infarction. Am J Cardiol 1982;50:428436. 2. Shah PK, Pichler M, Berman DS, Maddahi J, Peter T, Singh BN, Swan HJC. Non-invasive identification of a hioh risk subset of patients with acute inferior myocardial infarction. Am J Car&o1 1960;46:915-921. 3. Goldberg HL, Borer JS, Jacobstein JG, Kluger JI Schektt SS, Alonso DR. Anterior S-T segment depression in acute inferror myocardial infarction: indicator of posterolateral infarction. Am J Cardiol 1981;48: 1009-1015. 4. Gibson RS, Crampton RS, Watson DD, Taylor GJ, Carabello BA, Holt ND, Belier GA. Precordial ST-segment depression during acute inferior myocardial infarction: clinical, scintiaraohic and anaioaraohic correlations. - Circulation 1982;66:732-741. 5. Gelman JS, Saltups A. Precordial ST depression in patients with inferior mvocardial infarction: clinical imolications. Br Heart J 1982:48:5606. Ong L, Valdellon B, Coromilas J, Brody R, Reiser P, Morrison J. Precordial ST segment depression in inferior myocardial infarction. Evaluation by quantitative thallium201 scintigraphy and technetium-99m ventriculography. Am J Cardiol 1963;51:734-739. 7. Schuster EH, Bulkley BH. Early post-infarction angina. lschemia at a distance and ischemia in the infarct zone. N Engl J Med 1981;305:11011105. 6. Salcedo JR, Baird MG, Chambers RJ, Beanlands DS. Significance of re-

14. 15. 16. 17. 16. 19.

20.

THE AMERICAN

JOURNAL

OF CARDIOLOGY

Volume 55

329

ciprocal S-T segment depression in anterior precordial leads in acute inferior myocardial infarction: concomitant left anterior descending coronary artery disease? Am J Cardiol 1981;48:1003-1008. Roubin GS, Shen WF, Nicholson M, Dunn RF, Kelly DT, Harris PJ. Anterolateral ST segment depression in acute inferior myocardial infarction: angiographic and clinical implications. Am Heart J 1984;107:11771182. Little WC, Rogers EW, Sodums MT. Mechanism of anterior ST segment depression during acute inferior mvocardial infarction: observations durina coronary thrombolysis. Ann Intern Med 1984;100:226-229. Rosati RA. McNeer JF. Starmer CF. Mittler BS. Morris JJ. Wallace AG. A new information s&tern for clinical practice. Arch Intern Med 1975; 135:1017-1024. . ldeker RE, Wagner GS, Ruth WK: Alonso DR, Bishop SP, Bloor CM, Fallon JT, Gottlieb GJ, Hackel DB, Phdlips HR, Reimer KA, Roark SF, Rogers WJ, Savage RM, White RD, Selvester RH. Evaluation of a QRS scoring system for estimating myocardial infarct size. II. Correlation with quantitative anatomic findinas for anterior infarcts. Am J Cardiol 1982;49:16041614. Roark SF, ldeker RE, Wagner GS, Alonso DR, Bishop SP, Bloor CM, Bramlet DA. Edwards JE. Fallon JT. Gottlieb GJ. Hackel DB. Phillios HR. Reimer KA,‘Rogers WJ, Ruth WK, Savage RM? White RD, Selvesier RH: Evaluation of a QRS scoring system for estimatrng myocardial infarct size. III. Correlation with quantitative anatomic findings for inferior infarcts. Am J Cardiol 1983;51:382-389. Lehmann EL. Nonparametrics: Statistical Methods Based on Ranks. San Francisco: Holden-Day, 1975:379. Breslow N. Covariance analysis of censored survival data. Biometrics 1974;30:89-99. Cox DR. Regression models and life tables. J Royal Stat Sot (Series B) 1972;34:187-202. Severance HW, Morris KG, Wagner GS. Criteria for early discharge after acute myocardial infarction. Validation in a community hospital. Arch Intern Med 1982;142:39-41. Nasmfth J, Marpole D, Rahal D, Homan J, Stewart S, Sniderman A. Clinical outcomes after inferior myocardial infarction. Ann Intern Med 1982;96: 22-26. Wasserman AG, Ross AM, Bogaty D, Richardson DW, Hutchinson RG, Rios JC. Anterior ST segment depression during acute inferior myocardial infarction: evidence for the reciprocal change theory. Am Heart J 1983; 106:516-520. Billadello JJ, Smith JL, Ludbrook PA, Tiefenbrunn AJ, Jaffe AS, Sobel BE, Geltman EM. Implications of “reciprocal” ST segment depression associated with acute myocardial infarction identified by positron tomography. JACC 1983;2:616-624.