We read the results of the DIAD study (1) with great interest and congratulate all of the investigators. However, the study leaves some questions unanswered and raises others.
1) Since there is no nondiabetic control group with similar baseline characteristics, we don’t know whether the reported prevalence of silent ischemia (22%) is higher than in nondiabetic patients. Such a comparison is necessary to determine how many scans are needed to diagnose one addional silent ischemia in asymptomatic diabetic patients. Therefore, the clinician is not assisted in determining whether performing myocardial perfusion imaging (MPI) in asymptomatic diabetic patients has any incremental diagnostic and prognostic value.
2) Since patients are asymptomatic, it is unclear whether performing coronary angiogram (its cost, risks, and subsequent revascularization interventions) is cost-effective and/or will improve clinical outcomes.
3) The following factors may overestimate the prevalence of silent ischemia. i) Reported silent ischemia (22%; however, only 15.9% had abnormal MPI). Eight subjects (1.9%) had left ventricular dysfunction, which may not indicate silent ischemia. Categorizing transient ischemic dilatation and/or left ventricular dysfunction without perfusion defect as silent ischemia is also debatable. ii) Image analysis. Subjects were obese (BMI 31.1 ± 6.3 kg/m2), and 47% were women. It is not clear whether correction software for scatter, attenuation, and motion were used in the image interpretation. iii) Vasodilator stress testing using intravenous infusion of adenosine with simultaneous very-low-level treadmill exercise. The authors provide a rationale for this mode of stress test. However, it is of concern that asymptomatic, relatively young patients (60.7 ± 6.8 years) were not asked to perform an exercise test. False-positive scan rates are higher with vasodilator stress tests because myocardial images are more difficult to interpret due to relatively less myocardial uptake and increased gastrointestinal uptake. How generalizable are these findings in patients undergoing treadmill stress test with SPECT (single-photon emission–computed tomography) MPI? iv) Nine patients with transient ST-segment depression. It is not evident from the report whether these were purely adenosine or adenosine with exercise. The significance of electrocardiogram changes in vasodilator stress test with exercise is somewhat uncertain. v) Since there are no angiographic confirmations of abnormal scans, the sensitivity and specificity of positive scans are unclear.
4) Male sex as predictor of abnormal test. Despite the fact that the authors performed multivariate analyses, it would be helpful to know whether men and women differed on baseline characteristics.
In reference to issue 1, on retrospective analyses of our 231 MPIs in an inner-city hospital, the prevalence of abnormal MPIs among asymptomatic diabetic and nondiabetic patients was similar (20%) (2).
In view of all the above, a large-scale, multicenter, prospective study of asymptomatic diabetic and nondiabetic patients is warranted.
