Prognosis of silent coronary artery disease in diabetic patients

There has been a steady increase in the prevalence of type 2 diabetes. The disease correlates with a significantly greater risk of illness and death from cardiovascular disease. In patients with diabetes, 80% of deaths are related to atherosclerosis.1 Men with diabetes have a cardiovascular mortality rate two times greater than nondiabetic men, and diabetic women have a cardiovascular mortality rate four times greater than nondiabetic women.2 Furthermore, coronary artery disease (CAD) is often silent in diabetic patients.3 We have recently shown that, in patients with moderate to severe ischemia, more lives per year are saved with revascularization in diabetic patients compared with nondiabetic patients.4 Therefore, identifying CAD can have an important effect on the prognosis of asymptomatic diabetic patients.

To assess risk and diagnose patients with possible or known CAD, myocardial perfusion single photon emission computed tomography (SPECT) has been used extensively.5-7 Few studies have been done, however, on the relationship between symptoms in diabetic patients and the findings on myocardial perfusion SPECT testing. We compared the prognosis of CAD in symptomatic versus asymptomatic patients without a prior history of CAD.8

Patients and methods

The prognostic study cohort comprised 1,430 consecutive diabetic patients, all of whom received rest thallium-201/stress technetium-99m sestamibi myocardial perfusion SPECT. Standard protocols for stress testing and imaging were used.

Results

Patients were followed up for a median of 2 years. During the follow-up period, 98 cardiac events (nonfatal myocardial infarctions [MIs] or cardiac death) occurred, which resulted in a 3.1% annual cardiac event rate. Men had an event rate of 2.9%, and the event rate for women was similar, at 3.3%. With respect to risk and presenting symptoms, annual cardiac event rates were 2.2% for asymptomatic patients, 3.2% for patients with angina, and 7.7% for patients with shortness of breath. The event rates were markedly greater for patients with shortness of breath than for asymptomatic patients (P < .001) and those with angina (P = .002), reflecting in part the fact that patients with shortness of breath had a significantly lower ejection fraction. The event rates for asymptomatic patients and patients with angina were statistically the same.

Patients with abnormal myocardial perfusion SPECT results had a substantially higher annual event rate (5.4%) compared with patients with normal myocardial perfusion SPECT results (1.9%), including patients with angina, asymptomatic patients, and patients with shortness of breath (figure 1). Of note, annual event rates among patients with normal myocardial perfusion SPECT findings were similar for patients with angina, asymptomatic patients, and patients with shortness of breath. Among patients with abnormal myocardial perfusion SPECT results, however, patients with shortness of breath had significantly higher cardiac event rates than patients with angina (P = .008) and asymptomatic patients (P < .001); the outcomes for asymptomatic patients and patients with angina were similar. Also in patients with abnormal myocardial perfusion SPECT results, survival was better for patients with angina and asymptomatic patients than for patients with shortness of breath (figure 2).

Significant univariate predictors of cardiac events included the extent and severity of scarring, evidence of ischemia on myocardial perfusion SPECT, pharmacologic stress, age, hypertension, and shortness of breath as the presenting symptom. Based on survival modeling using a Cox proportional hazards model, extent of ischemia, extent of scarring, hypertension, shortness of breath, and age were most predictive of events and, added incrementally to each other. Importantly, angina was not a significant predictor of critical events.

Discussion

CAD represents the leading cause of death in diabetic patients. Silent myocardial ischemia occurs more often in diabetic than in nondiabetic patients. Our data show that, in a diabetic patient population referred for CAD evaluation, 39% of patients had myocardial perfusion SPECT evidence of silent CAD. The rate

of abnormal myocardial perfusion SPECT results did not differ from that of patients with angina or angina-like chest pain. However, the occurrence of CAD was much greater in patients with a history of shortness of breath, as shown by abnormal myocardial perfusion SPECT results. The fact that the absence of angina does not necessarily mean that there is no CAD in diabetic patients is highlighted by these results.

Among those who had a normal myocardial perfusion SPECT scan, the outcome was the same for asymptomatic patients, those with angina, and those with shortness of breath in our study. For diabetic patients with evidence of CAD, the outcome was no different for patients with and without angina. In patients with shortness of breath, however, the outcome was much worse. In part, this was probably because these patients had a lower left ventricular ejection fraction than asymptomatic patients and patients with angina, as well as evidence of more previous silent MIs as shown on myocardial perfusion SPECT scanning.

In the first large prospective trial examining silent CAD in diabetic patients, the Detection of Ischemia

in Asymptomatic Diabetics (DIAD) trial, 27% of patients in a low-risk diabetic population had myocardial perfusion SPECT evidence of CAD.9 Eighteen percent of patients had silent ischemia, 3% had evidence of scarring, and 6% were identified as being at high risk. The rates of silent CAD in that study were lower than in our study, which included a higher risk group and older diabetic patients. Interestingly, in the DIAD study, the presence of autonomic cardiac neuropathy was the only independent predictor of an abnormal myocardial perfusion SPECT scan. In the presence of myocardial ischemia, diabetic patients have been shown in several studies to report angina less often than nondiabetic patients,10 and the only symptom of myocardial ischemia may be shortness of breath.11 The findings of these previous reports and our study may also be related to an underlying autonomic cardiac neuropathy.

The importance of looking for silent ischemia in diabetic patients is illustrated in another recent study from our laboratory.4 As noted earlier, we reported that diabetic patients manifest a greater survival benefit with revascularization in the setting of extensive reversible ischemia than nondiabetic patients. This enhanced survival benefit is greater in women than in men.

Another important aspect of the present study is the fact that, besides the objective evidence of CAD, hypertension was identified as an independent predictor of outcome in diabetic patients. This finding is in agreement with earlier studies that have shown the importance of strictly controlling hypertension in patients with diabetes.12,13

The American Diabetes Association published a consensus development conference report in 1998 containing recommendations on how to screen asymptomatic diabetic pa-tients with respect to CAD.14 Rather than being based on evidence, however, the guidelines were based on the clinical opinions of the authorities who wrote them. There is a clear need for evidence-based guidelines for the early detection of CAD in patients with diabetes mellitus.15 The results of the current study provide additional data and suggest a place for nuclear testing in asymptomatic diabetic patients. The myocardial perfusion SPECT results will be important for decision making and choosing the appropriate treatment strategies. There are few studies on the treatment of diabetic patients with early asymptomatic CAD. In the Asymptomatic Cardiac Ischemia Pilot study, however, revascularization was shown to markedly reduce MI, cardiac hospitalizations, and death in patients with silent ischemia compared with patients treated with medication.16

Conclusions

In our study, silent CAD was commonly found in asymptomatic diabetic patients referred for CAD assessment. It occurred as frequently as in diabetic patients with angina. Accordingly, the annual critical event rate was similarly and strongly dependent on objective evidence of CAD (as demonstrated on myocardial perfusion SPECT scans). Partly because of previous silent MIs, as indicated by a poor left ventricular ejection fraction and larger regions of scarring, patients with shortness of breath had a significantly worse prognosis than asymptomatic patients and patients with angina. These findings point to the fact that objective evidence of CAD shown on myocardial perfusion SPECT can be important in predicting the prognosis and enhancing the decision-making process for asymptomatic diabetic patients. n