Prospective Temporal Analysis of the Onset of Preinfarction Angina Versus Outcome
An Ancillary Study in TIMI-9B
Background—The timing of onset of angina before myocardial infarction in relation to outcome is unknown.
Methods and Results—We prospectively determined the importance of the time of onset of preinfarction angina in relation to 30-day outcomes in the TIMI-9B study from standardized forms. Of the 3002 patients entered into the study, 425 reported angina before their myocardial infarction. Patients with angina onset within 24 hours of infarction had a lower 30-day cardiac event rate (mortality, recurrent myocardial infarction, heart failure, or shock) at 4% than those with onset of angina >24 hours (17%; P=.030). A history of any angina alone was not associated with reduced event rate. Peak creatine kinase levels tended to be lower in the group with angina within 24 hours. These benefits were not due to higher rates of use of antianginal medicines or aspirin and were not a consequence of differences in baseline characteristics or disease states (hypertension, hypercholesterolemia) among subgroups.
Conclusions—These temporal observations are consistent with the concept of preconditioning by preinfarction angina but do not rule out other mechanisms.
Preinfarction angina pectoris has been suggested in some studies to have beneficial effects on myocardial infarction (MI).1 2 3 4 5 6 The exact mechanism of this protective effect is not known, but it may include ischemic preconditioning. If preconditioning plays a role, then anginal episodes closest to the time of infarction should have the most benefit. However, there are few data available regarding the relation between the time of onset of angina and outcome in MI. Moreover, many of the previous analyses of angina and MI have been retrospective rather than prospective.1 2 6 Thus our objective was to assess prospectively the relation between the time of onset of angina before MI and the incidence of primary end points of the trial (death, recurrent MI, congestive heart failure, or shock) during the first 30 days after MI in patients enrolled in the TIMI-9B study. Our primary hypothesis was that patients in whom the time between onset of angina and MI was within 24 hours would have different outcomes compared with patients who had onset of angina >24 hours before infarction.
The study included 3002 patients with acute MI who were randomized in the TIMI-9B study, a prospective trial designed to assess the effect of hirudin versus heparin as adjunctive therapy to thrombolysis.7 This study was an ancillary study to TIMI-9B that was designed prospectively. Details regarding the basic TIMI-9B protocol and inclusion and exclusion criteria have been published elsewhere.7 Inclusion criteria for MI were ischemic discomfort ≥30 minutes in duration and ST-segment elevation ≥0.1 mV in two contiguous ECG leads or new or presumably new left bundle branch block. The presence or absence and date of onset of angina before MI was noted on standardized forms. These forms were completed by trained research coordinators, and independent monitoring of the data was carried out in all patients. Instructions were provided to all investigators regarding definition of angina: “Angina pectoris is a discomfort in the chest with or without radiation to the arms, neck, or back, which is caused by myocardial ischemia and can be associated with disturbances in myocardial function, but not necrosis. This is an important distinction because pain of infarction should not be classified as angina. In most cases angina is precipitated by exertion and is generally relieved by rest.” In general, angina duration is <20 to 30 minutes. Ischemic discomfort ≥30 minutes in duration was in fact one aspect of eligibility criteria for inclusion as MI, not angina in this protocol. “Angina is a constellation of sensations with chest discomfort often described as ‘constricting, crushing, heaviness, or squeezing.’ It can be a vague numbness or pressure sensation. In some patients ‘angina equivalents’ are present such as shortness of breath, fatigue, and belching. When evaluating a patient’s symptoms and daily limitations due to ischemia, all features described by the patient must be taken into consideration.”
The following cumulative major cardiac events within 30 days of admission for hospitalization were noted: death, recurrent MI, new-onset (>4 hours after randomization) severe congestive heart failure (rales over more than half the lung fields and pulmonary congestion on chest radiograph), or new-onset (>4 hours after randomization) cardiogenic shock (end-organ hypoperfusion, systolic pressure <90 mm Hg without inotropic or intra-aortic balloon support or >90 mm Hg requiring inotropic or intra-aortic balloon support in the setting of adequate volume expansion). The definition of recurrent MI was based on ECG and enzyme criteria.7 Creatine kinase (CK) levels were obtained at baseline, every 8 hours ×2 on the first day, and then daily. Peak CK units were also recorded. Event rate and CK release were compared among patients with time between onset of angina and MI ≤24 hours, 48 hours, 72 hours, 1 week, 1 month, or >1 month before infarction, and patients with no preinfarction angina. The data were further analyzed to compare the outcomes between the patients with preinfarction angina ≤24 hours versus those with angina >24 hours. The patient characteristics at baseline were compared between those with angina ≤24 hours and those with angina >24 hours. For all analyses, the χ2 test was applied for categorical variables and one-way ANOVA for continuous variables. Effects of potential confounding factors (including previous disease and use of antianginal medications) were evaluated with the Cochran-Mantel-Haenszel test and Breslow-Day test. All statistics were performed by Corning Besselar Covance at Princeton, NJ.
Of the 3002 patients enrolled in TIMI-9B, 425 reported angina before their MI and 2577 did not.
Table 1⇓ shows the event rates in relation to the time between onset of reported angina and MI. There was a difference in the event rate among these multiple categories (P=.012, by χ2). Patients in whom the time between onset of reported angina and MI was within 24 hours (≤24 hours) had lower event rates (4%) versus those in whom this duration was >24 hours (17%; P=.03).
Table 2⇓ shows the comparison of maximum CK values in relation to the duration of reported onset of angina before MI. Patients with angina ≤24 hours had a nonsignificant trend toward lower maximum CK values (mean±SE, 1133±147) compared with those with angina >24 hours (mean±SE, 1416±55 Units, P=.09). In addition, patients in whom the time between onset of angina and MI was within 24 hours had lower maximum CK values (1133±147) compared with those without angina (1570±21; P<.01). Patients with any history of angina had lower maximum total values (1386±52) compared with those without any history of angina (1570±21; P<.01). The duration between the angina and the onset of infarction did not correlate with CK values or outcomes when the duration was treated as a continuous rather than categorical variable. However, the duration of 24 hours may be a key point that divides patients into these two heterogeneous groups.
Table 3⇓ shows some of the characteristics of patients in whom the time between onset of angina and MI was within 24 hours, in whom this duration was >24 hours, in whom there was any history of angina and in whom there was no history of angina. Patients with a history of hypertension were more likely to have angina ≤24 hours (P=.04 versus those with angina>24 hours). There was a nonsignificant trend toward a higher frequency of hypercholesterolemia (P=.06) among patients with onset of angina ≤24 hours versus >24 hours. Patients with angina >24 hours had a nonsignificant trend toward more aspirin use than patients with angina ≤24 hours (P=.06). The Breslow-Day test was used to examine homogeneity of differences in event rates across levels of each confounding factor including use of drugs. None of the drugs used and previous diseases affected the association between the cardiac event rates and time of onset of preinfarction angina.
The results of this prospective analysis suggest that the benefit of preinfarction angina on clinical event in the TIMI-9B study are only manifest when the time between onset of angina and MI are within 24 hours. This time frame is consistent with (but not proof of) the concept of cardioprotection by either “classic preconditioning” and/or “delayed or late preconditioning.”8 9 With classic preconditioning, the ischemic preconditioning episodes occur within a few hours of infarction; with delayed or late preconditioning the onset of ischemic preconditioning occurs 12 to 24 hours after infarction. Some studies have even suggested that the myocardium can be protected if the preconditioning occurs 48 to 72 hours before the infarction; although the results of the present study suggest that in humans, brief ischemia needs to occur within 24 hours to be protective. However, other potential mechanisms besides preconditioning may play a role in the findings of this study.
For example, preinfarction angina may initiate the development of new collateral vessels, which could be protective. Although previous studies found no difference in the presence of large epicardial vessels in patients with antecedent angina,1 2 3 4 the contribution of small intramural or microvascular collaterals not detected by coronary angiography cannot be excluded. However, in the present study, it is unlikely that the beneficial effects of preinfarction angina occurring within 24 hours before infarction are due to new collateral growth because new collaterals require >24 hours to develop. Another potential explanation is that angina ≤24 hours before MI is associated with better and earlier thrombolysis.10 Because angiography was not a required part of the current protocol, we do not have systematic angiographic data. However, as part of the TIMI-4 study, we did observe that of 79 patients who had angina within 48 hours of infarction, 46% had TIMI grade 3 flow on 60-minute angiograms, as compared with 34% of 140 patients with no angina. Patients with preinfarction angina were less likely to have persistent occlusion (TIMI grade 0) at 60 minutes (19%) compared with those without angina (27%).11 In the current study (TIMI-9B) we did not assess either collateral perfusion or time to angiographic reperfusion, and these clearly are limitations. However, we have recently observed that brief antecedent ischemia attenuates platelet-mediated thrombosis and preserves vessel patency in damaged and stenotic coronary arteries.12 Finally, one could argue that the use of antianginal medicines and aspirin might account for the benefit of preinfarction angina. However, as indicated in Table 2⇑, there were no significant differences in drug use among the cohorts, and the Breslow-Day test showed that the drugs used did not affect the association between onset time of angina and cardiac events.
This study is, in some respects, similar to previous ones investigating the effect of preinfarction angina on outcome. Like others, preinfarction angina was beneficial1 2 3 4 5 6 12 in that it was associated with lower CK values and better outcomes. However, in this study, our goal was to assess the importance of timing of onset of angina. In our study, preinfarction angina was only beneficial for event rate when the time between onset of angina and MI was within 24 hours; a history of any angina alone was not associated with a lower event rate. Why this benefit was observed only when angina occurred within 24 hours while other studies have suggested that any preinfarction angina is protective is not clear. Certainly the timing suggests a preconditioning mechanism. The fact that any angina was associated with a lower peak CK but not event rate suggests that a preconditioning effect on event rate may be related to more than just extent of necrosis and could involve effects on factors such as left ventricular function and vascular stability. Also, one recent preliminary report13 suggested that only prodromal angina within 24 hours of the infarction was associated with improved long-term prognosis. There are other important differences between this and previous reports. First and most notably, this is one of the few prospective studies that has investigated preinfarction angina in a large multicenter trial looking at separate time periods before infarction. Second, the percentage of patients with preinfarction angina in the present study was lower than that reported in our own previous analysis of the TIMI-4 study.1 The reasons for this difference in TIMI-9 compared with TIMI-4 is not known but may represent differences in the populations studied. Finally, in TIMI-4, any preinfarction angina appeared to confer protection, while in this study angina within 24 hours was most protective compared with angina before 24 hours. We did not, however, perform a detailed time between onset of angina and MI analysis in the TIMI-4 study. Thus the novel observation of the present prospective temporal analysis is that onset of angina within 24 hours of MI is associated with a lower 30-day cardiac event rate and trend toward lower peak CK compared with angina before 24 hours.
One limitation of this type of study is that it is not possible to obtain controlled data regarding how much silent ischemia might have occurred before infarction. None of the patients were or could have been fitted with ambulatory ECG monitors before their MI. It is certainly conceivable that silent ischemia might have had a potential additive effect on preconditioning. A recent study by Sandhu et al14 performed in a rabbit MI model showed that three cycles of ischemic preconditioning were more effective than one cycle in reducing infarct size. It is possible that some patients in our study had silent ischemia, which could have further preconditioned the myocardium. However, it is likely that those patients with some angina were more likely to also have silent ischemia.
Why patients who had onset of preinfarction angina within 24 hours also had more hypertension and a trend toward more hypercholesterolemia is unknown. However, it is interesting that despite a higher incidence of these risk factors, the patients still had better early outcomes than patients without preinfarction angina. There was a trend, although not statistically significant, for patients with angina >24 hours to have more prior MI. Although we cannot rule out the possibly that this contributed to a worse outcome, in the TIMI-4 study patients with previous MI were more likely to have prior angina, which was associated with a better outcome.1
One potential clinical implication is that if infarctions are smaller in patients who have angina within 24 hours of infarction, the window of opportunity for thrombolysis may be wider. That is, patients with this preinfarction angina may potentially benefit from thrombolysis even if instituted beyond 12 hours. This possibility remains to be tested in clinical trials. However, it probably is worthwhile to question patients regarding the time of onset of preinfarction angina in order to provide clues as to early outcomes.
We are indebted to Kisook Yoo, PhD, Qi Liu, and Juan T. Torres of Covance and Corning Besselar for statistical analyses and Cathy Davisson for preparation of the manuscript.
The complete list of TIMI-9B Investigators is found in the Appendix of reference 7.
- Received June 27, 1997.
- Revision received November 20, 1997.
- Accepted November 24, 1997.
- Copyright © 1998 by American Heart Association
Kloner RA, Shook T, Przyklenk K, Davis V, Junio L, Matthews RV, Burstein S, Gibson M, Poole WK, Cannon CP, McCabe CH, Braunwald E, for TIMI 4 Investigators. Previous angina alters in-hospital outcome in TIMI 4: a clinical correlate to preconditioning? Circulation. 1995;91:37–47.
Ottani F, Galvani M, Ferrini D, Sorbello F, Limonetti P, Pantoli D, Rusticali F. Prodromal angina limits infarct size: a role for ischemic preconditioning. Circulation. 1995;21:291–297.
Antman EM, for the TIMI 9B Investigators. Hirudin in acute myocardial infarction: thrombolysis and thrombin inhibition in myocardial infarction (TIMI) 9B Trial. Circulation. 1996;94:911–921.
Hata K, Kloner RA, Przyklenk K. Brief antecedent ischemia attenuates platelet-mediated thrombosis in damaged and stenotic canine coronary arteries. Circulation. 1996;94(suppl I):I-558. Abstract.
Ishihara M, Sato H, Kawagol T, Sakai K. Implication of prodromal angina pectoris on long-term prognosis after acute myocardial infarction. Circulation. 1996;94(suppl I):I-611–I-612.
Sandhu R, Diaz R, Mao GD, Wilson GJ. Ischemic preconditioning: differences in protection and susceptibility to blockade with single-cycle versus multicycle transient ischemia. Circulation. 1997;96:984–995.