The DIAD study was designed to investigate the prevalence and predictors of silent ischemia, specifically in asymptomatic patients with type 2 diabetes (1), who often present with advanced coronary artery disease (CAD) without prior angina. The DIAD study did not compare the prevalence of asymptomatic CAD in diabetic and nondiabetic patients. Although the relative prevalence of asymptomatic CAD in nondiabetic patients otherwise well-matched for cardiac risk factors is of interest, it will require a much larger study.
Bhalodkar and Blum (2) express concern that the prevalence of silent ischemia in DIAD might be overestimated. We believe this to be unlikely. All single-photon emission–computed tomography (SPECT) studies were interpreted by consensus of a panel of well-recognized experts in nuclear cardiology. Attenuation correction was not applied, and images were read visually along with quantitative analysis. By design, all images with possible attenuation artifacts were interpreted as normal. We would agree that resting left ventricular dysfunction and transient ventricular dilation without perfusion defects may not be due to macrovascular CAD, but these were infrequent findings.
The specificity of SPECT Sestamibi imaging abnormalities in asymptomatic patients with diabetes has not been formally evaluated. However, we previously established a normalcy rate of 97% in normal subjects using the same quantitative analysis method. Importantly, our analysis of the predictors of ischemia focused on patients with moderate or large reversible SPECT defects, which occur rarely in the absence of significant CAD.
Since many patients with diabetes are unable to exercise adequately, vasodilator stress was chosen to ensure that the study was broadly inclusive. Exclusion of patients unable to exercise would select a lower risk and nonrepresentative study group. The literature does not indicate that vasodilator stress has a higher false-positive rate than exercise stress, and adenosine Sestamibi images are not more difficult to interpret than exercise images in experienced laboratories. By study design, poor-quality studies were excluded; however, only 1 of 523 SPECT studies was deemed of poor quality.
The adverse prognostic significance of ST-segment depression with adenosine is well documented in patients referred for evaluation (3,4). Most nuclear cardiologists consider this finding to indicate an increased risk for future cardiovascular events. Of patients with ST-segment depression, 5 of 9 patients with perfusion defects and 13 of 21 patients without perfusion defects performed low-level exercise (P = NS). The second phase of DIAD (to be completed in September 2007) will define the prognostic significance of adenosine-induced ST-depression in asymptomatic patients with diabetes.
We agree that, currently, there are insufficient data to support routine angiography/intervention in all patients with abnormal screening SPECT studies. Physicians who screen patients must use their own judgment in determining individual management at the current time. The follow-up phase of DIAD will lend insight into which patients are at highest risk for future cardiovascular events. However, an abnormal perfusion study should trigger intensive risk factor intervention that is otherwise not commonly achieved, even when patients are treated in the diabetes clinic.
With regard to the unpublished observations mentioned in the letter, caution is warranted in interpreting retrospective data from a referred patient population because of selection bias. We appreciate the interest in DIAD.
