Gated SPECT after myocardial infarction: New techniques/old principles

G technetium-99m sestamibi single-photon emission computed tomographic (SPECT) imaging has grown rapidly over the past few years, and is now widely used for the assessment of both regional and global left ventricular function. This technique has been well validated.1,2 The incremental value of ejection fraction (EF) measurements that it provides has been clearly demonstrated in patients with chronic chest pain and suspected coronary artery disease.3 This issue of the journal contains one of the first studies of the clinical use of this technique early after myocardial infarction.4 This single-center study has several limitations. The number of patients is small (124), the number of hard events is also small (10 deaths and 10 myocardial infarctions), and many of the potentially eligible patients were not included because they underwent early revascularization after stress-gated SPECT sestamibi imaging. Because of the small number of events, the authors analyzed the combination of deaths and nonfatal myocardial infarctions. Most previous outcome studies have focused exclusively on mortality.5–8 The pathophysiology of nonfatal myocardial infarction may be considerably different.9 Despite these limitations, Kroll et al4 demonstrate that a history of prior myocardial infarction, low exercise capacity, and an EF ,40% by gated SPECT imaging are all independent predictors of hard events during the 18.9 months after myocardial infarction. These novel results for this new technique reinforce several “old” established principles. Clinical characteristics are important predictors of subsequent patient outcome. Despite the small number of events in this study, both diabetes mellitus and history of myocardial infarction were significant univariate predictors of subsequent hard events. On multivariate analysis, after adjustment for both exercise capacity and EF, a history of myocardial infarction remained highly significant with a relative risk of 5.3. Multiple previous studies have demonstrated the prognostic importance of clinical characteristics.8,10 A rigorous multivariate analysis of 30-day mortality in 41,021 patients from the Global Utilization of Streptokinase and t-PA for Occluded Coronary Arteries trial found that both history of myocardial infarction and diabetes mellitus were highly significant independent predictors with risk ratios of between 1.5 and 2.0.8 However, age and measures of acute hemodynamics (systolic blood pressure, Killip class, and heart rate) were far more powerful predictors.8 Age alone accounted for 31% of the mortality, compared with only 2.8% for history of myocardial infarction and 2.5% for hypertension, diabetes mellitus, and smoking combined. The failure of age to achieve clinical significance in this study may reflect its small size, the play of chance, or a restricted distribution of age in the study group. Because clinical characteristics are readily available at very little cost, they should ideally be “forced into” any analysis of patient outcome after myocardial infarction. The analysis performed by Kroll et al4 is similar to many other analyses in the literature that have required clinical characteristics to “compete” in multivariate models with the results of more expensive tests. Although such an analysis does not properly reflect clinical decision making,11 it is often necessary in small studies to avoid spurious results due to model overfitting. The significance of a history of myocardial infarction as an independent predictor of hard events in this study reinforces the importance of clinical characteristics as predictors of outcome. Exercise capacity is a major predictor of outcome after myocardial infarction. The authors chose to analyze both exercise and dipyridamole stress studies by creating a novel “exercise score” that combined patients who required dipyridamole with those who could not exercise to a level of 4 METs on a Bruce protocol. This dichotomous variable was highly significant on both univariate and multivariate analyses, with a relative risk of 6.84 after adjustment for prior myocardial infarction and left ventricular function. Prior studies have shown that the highest risk subset of patients following myocardial infarction are those patients who are unable to exercise.12 In the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico (GISSI-2) study,12 patients who were unable to exercise had a 7% 1-year mortality compared with 1.0% of those patients who were able to exercise.12 In patients who are able to exercise, exercise capacity and systolic blood pressure response are much more important predictors of outcome that STsegment depression, which is often the focus of greater clinical attention. In a meta-analysis of 15,613 patients, Shaw et al13 demonstrated that low exercise capacity and abnormal systolic blood pressure response were both associated with risk ratios of approximately 4.0, compared with a risk ratio of only 1.4 for ST-segment depression. The prognostic importance of exercise capacity has been reiterated in recent studies in outpatient populations with chronic symptoms.14,15 It is not widely appreciated that exercise capacity has prognostic value that is incremental to From the Mayo Clinic, Rochester, Minnesota; and North Texas Heart Center, Presbyterian Hospital of Dallas, Dallas, Texas. Manuscript received and accepted August 29, 2000. Address for reprints: Darryl L. Kawalsky, MD, North Texas Heart Center, 8440 Walnut Hill Lane, Suite 700, Dallas, Texas 75231-3824.