Increased Carotid Intima-Media Thickness as a Predictor of the Presence and Extent of Abnormal Myocardial Perfusion in Type 2 Diabetes

Prospectively, 98 consecutive asymptomatic patients with type 2 diabetes (1) were recruited from a routine outpatient diabetes clinic and referred for cardiovascular risk stratification. Asymptomatic status was confirmed using the Rose questionnaire (2). All patients underwent myocardial perfusion imaging by single-photon emission computed tomography (SPECT) and CIMT assessment.

Myocardial perfusion imaging was performed using electrocardiogram-gated SPECT with ^99mTc-sestamibi during pharmacological stress and rest, according to protocols described previously (3).

Using a 17-segment model tracer, uptake in each segment was evaluated by two observers in consensus by use of a five-point scoring system (4). The total segmental score during stress was used to determine the extent of abnormal perfusion as reflected by the summed stress score (SSS). Abnormal perfusion was defined as SSS ≥3 and severely abnormal perfusion as SSS ≥8.

CIMT was assessed using high-resolution B-mode ultrasound with a 10-MHz linear transducer, with an automatic boundary detection system (Art Lab-Esaote-Picus, Genova, Italy). Measurements were performed by an experienced sonographer blinded to clinical information (5). Mean CIMT was assessed throughout 10-mm segments, at four angles, across the far wall of the right and left common carotid artery (CCA). The average of the mean CIMT values of the four segments was calculated to determine the mean right and left CIMT per patient.

CIMT values ≥75th percentile (per age and sex category) are defined as increased, indicating elevated cardiovascular risk (5). In the current study, the mean CIMT was compared with reference values from the Multi-Ethnic Study of Atherosclerosis (5). Patients were thereby stratified as having normal CIMT (CIMT <75th percentile) or increased CIMT (CIMT ≥75th percentile in at least one CCA).

First, average SSS and standard deviations were calculated in patients with normal or increased CIMT. The independent T test was used to assess the difference in mean SSS between the two groups.

Thereafter, univariate analysis of baseline characteristics including age, positive family history of CAD, smoking, hypertension, hypercholesterolemia, BMI, nephropathy (urine albumin/creatinine ≥3.5 mg/mmol), fasting glucose, glycated hemoglobin (by chromatography) (6), retinopathy, peripheral arterial disease, and increased CIMT was performed to identify potential predictors of the extent of abnormal perfusion (SSS). Subsequently, risk factors with a P value <0.05 were included in a linear multiple regression model to identify independent predictors of SSS. Finally, the prevalence of abnormal perfusion (SSS ≥ 3) and severely abnormal perfusion (SSS ≥ 8) was compared between patients with normal and increased CIMT.

Briefly, the mean age of the study population was 54 ± 11 years, with the majority of patients being male (n = 50, 51%). Mean SSS was 3.1 ± 4.2 in the total population. Overall, 34 patients (35%) showed abnormal perfusion (SSS ≥ 3), including severely abnormal perfusion (SSS ≥ 8) in 14 patients (14%). Average CIMT was 0.68 ± 0.12 mm. Comparison with reference values revealed normal CIMT in 60 patients (61%), while in the remaining 38 patients (39%) an increased CIMT value was observed in at least one CCA.

The mean SSS increased significantly from 1.2 ± 2.1 in patients with normal CIMT to 5.6 ± 4.6 in patients with increased CIMT (P < 0.001).

Age, smoking, hypertension, nephropathy, and increased CIMT were identified as potential predictors of SSS on SPECT, in a univariate regression model. Importantly, after adjustment for age, smoking, hypertension, and nephropathy in a multivariate model, increased CIMT remained a significant predictor of SSS (P < 0.001) (β = 4.41 [95% CI 3.05–5.76]).

Abnormal perfusion was present in 9% of patients with normal CIMT versus 75% of patients with increased CIMT (Fig. 1,A). Notably, prevalence of severely abnormal perfusion increased from 3% in patients with normal CIMT to 28% in those with increased CIMT (Fig. 1 B).

The prognostic value of SPECT imaging has been confirmed in diabetic patients (7). In particular, a favorable cardiovascular prognosis has been described in patients with normal myocardial perfusion, whereas significantly higher adverse event rates were observed in patients with severely abnormal perfusion (8). SPECT has therefore been proposed as a screening tool for identification of asymptomatic diabetic patients with obstructive CAD (9). However, considering the high global prevalence of type 2 diabetes, a broad screening strategy of all asymptomatic patients using SPECT perfusion imaging does not appear feasible or cost-effective (10). The American Heart Association/American Diabetes Association therefore initially suggested more aggressive medical treatment and assessment of CAD only in the presence of two additional risk factors (9). Nonetheless, baseline analysis of SPECT data in the Detection of Ischemia in Asymptomatic Diabetics Study demonstrated that a selection strategy based on a minimum of two additional risk factors underestimates the presence of abnormal perfusion in a large proportion of patients (41%) (11). Accordingly, the key question remains how asymptomatic diabetic patients with severe CAD should be identified from the general diabetic population.

Assessment of CIMT has been previously proposed for this purpose (12). Moreover, the truly noninvasive, inexpensive, and radiation-free nature of CIMT may represent an important advantage over other suggested screening techniques such as coronary calcium scoring (5). However, the relation of CIMT with CAD has not been fully established in asymptomatic diabetic patients. In the current study, increased CIMT was shown to be a strong predictor of the extent of abnormal perfusion and improved identification of patients with severely abnormal perfusion (28%). Normal CIMT values on the other hand were associated with a low risk for abnormal perfusion. Importantly, only few asymptomatic diabetic patients with normal CIMT values had severely abnormal perfusion (3%). However, it must be acknowledged that a nondiabetic control group was not available for comparison. Also, patients were referred from a diabetes clinic and may thus represent a more high-risk group than the general asymptomatic population with diabetes. Nevertheless, our findings suggest that initial risk stratification using CIMT may allow selective referral of asymptomatic patients with type 2 diabetes requiring further imaging and intensification of therapy, thereby improving patient outcome while maintaining cost-effectiveness.

J.J.B. received grants from Medtronic (Minneapolis, MN), Biotronik (Berlin, Germany), Boston Scientific (Natick, MA), Bristol-Myers Squibb Medical Imaging (North Billerica, MA), St. Jude Medical (St. Paul, MN), GE Healthcare (Buckinghamshire, U.K.), and Edwards Lifesciences (Irvine, CA). J.W.J. received grants from Biotronik (Berlin, Germany), Boston Scientific (Natick, MA), AstraZeneca (London, U.K.), Pfizer (New York, NY), and Merck Sharp and Dohme (Readington Township, NJ).

No other potential conflicts of interest relevant to this article were reported.

We thank M. Spaans and J. op 't Roodt for their skilled ultrasonographic analysis of the CIMT.