Lower-Limb Vascularization in Diabetic Patients: Assessment by thallium-201 scanning coupled with exercise myocardial scintigraphy

A total of 80 patients (44 men, 36 women) were prospectively recruited in the Department of Diabetology of Jean Verdier Hospital (Bondy, France). Mean age was 57.3 years (range 36–77). Four patients had type 1 diabetes and 76 patients had type 2 diabetes that was treated by diet and oral hypoglycemic drugs or insulin. The duration of disease was ≥15 years in the patients with type 1 diabetes and ≥5 years in the patients with type 2 diabetes. The mean duration of diabetes was 13 years (range 5–36). The criteria for inclusion were absence of myocardial infarction or angina pectoris, normal results of 12-lead resting electrocardiography (ECG), and absence of claudication. In seven patients, one foot pulse was not palpable. In these seven patients, an LL arterial ultrasound examination was performed to ensure that no PVD was present. Results were normal in four patients and showed slight irregularities of the femoral or tibial arteries in the three others. No carotid or femoral murmur was heard in any of the patients. None of the patients had a history of stroke. All patients had at least one additional cardiovascular risk factor: dyslipidemia (serum total cholesterol >6.5 mmol/l and/or triglycerides >2.3 mmol/l and/or lipid-lowering treatment), hypertension (blood pressure >140/80 mmHg or antihypertensive treatment), past or current use of tobacco (>10 pack-years of cigarette exposure), or urinary albumin excretion rate >30 mg/day.

None of the patients had congenital heart disease or cardiomyopathy, asthma or severe chronic bronchopathy, renal failure with plasma creatinine >250 μmol/l, thyroid disease, chronic alcohol consumption, or trauma/surgery involving the LL.

Presence of diabetic retinopathy was assessed by an experienced ophthalmologist through dilated pupils by direct ophthalmoscopy and fluorescein angiography. Diabetic retinopathy was diagnosed if at least one microaneurysm, hemorrhage, or exudate in one eye was observed. The condition was classified as nonproliferative in 21 patients (14 mild, 1 moderate, 6 very severe) and proliferative in 3 patients (20). The presence of nephropathy, defined by an albumin excretion rate >30 mg/day on at least two assessments, was diagnosed in 20 patients, including 4 patients with clinical albuminuria (>300 mg/day) (21). Evidence of neuropathy was obtained on the basis of at least one objective sign (including absence of both ankle tendon reflexes and impairment of LL touch and pin-prick sensation), whether or not the patient had symptoms (22). The baseline data of the patients are summarized in Table 1.

Stress ²⁰¹Tl myocardial scintigraphy was performed on all patients. Briefly, ECG was performed during graded exercise testing on a bicycle ergometer. The workload from an initial level of 30 W was increased by 30 W every 3 min. Cardiac rhythm was monitored continuously, 12-lead ECG was performed, and arterial blood pressure was measured at baseline and at the end of each work load. The results of the test were considered positive if angina pectoris occurred or if there was an ST-segment horizontal or downsloping depression ≥0.1 mV at 0.08 s beyond the J point, ventricular ectopic activity, atrioventricular blockade, or intraventricular bundle blockade. The results of exercise ECG were interpreted as negative if there were no ischemic changes during exercise and the patient achieved >85% of the maximum predicted heart rate. At the near-maximal peak of exercise, 1.5 MBq/kg ²⁰¹Tl was administered, and exercise was continued for 1 min more. Single-photon emission computed tomography (SPECT) imaging was performed within 5–10 min after the injection of ²⁰¹Tl, with the patient in the supine position, using a single-head, wide-angle-lens gamma camera (Sophycamera DSX Rectangular; SMV, Buc, France) and a low-energy, high-resolution collimator. Delayed images were obtained 4 ± 0.5 h later. The perfusion pattern in each vascular territory was assessed as normal or showing stable or reversible abnormalities (23). SMI was defined by positive results of ECG stress testing and/or exercise myocardial SPECT.

Patients were instructed to remain in the supine position for LL ²⁰¹Tl scanning in the same run after exercise myocardial SPECT imaging, <35 min after peak exercise. A single 512 × 512 pixel posterior pass was used with a scanning speed of 15 cm/min (11). Data were again stored on a magnetic disk linked to a SOPHY NXT/P workstation (SMV, Buc, France). Three irregular areas of interest were drawn using a trackball around the entire buttock, thigh, and calf on one side. Symmetrical areas of interest were automatically generated (Fig. 1). No correction was made for background activity. Total counts in each of the three levels were then determined by computer, and right-to-left ratios were calculated at each level. A ratio <0.9 or >1.1 at one or several levels defined an abnormal result of LL scintigraphy for the level considered, as reported previously (11). The perfusion defects were obvious in some patients, but counting was always performed.

For patients with SMI, cardiac catheterization was proposed. Left heart catheterization and coronary arteriography were performed by the percutaneous femoral approach. Coronary angiograms were analyzed by two experienced investigators. Patients were considered to have coronary artery disease if coronary stenosis was ≥50% in the left coronary artery or ≥70% in the left anterior descending artery, the circumflex artery, a well-developed marginal vessel, or the right coronary artery.

The following measurements were recorded: HbA_1c (integration of the electrophoretic peak; Hyrys Hydrasys, Sebia, France; normal value <6%), serum total cholesterol and triglycerides, HDL cholesterol (enzymatic colorimetry; Hitachi 912; Roche-Boehringer, Meylan, France), plasma creatinine (kinetic colorimetry; Kone, Helsinki, Finland), and the 24-h urinary albumin excretion rate (laser immunonephelometry; BN100; Dade-Behring, Paris la Défense, France) were measured.

The statistical analyses were performed using SPSS software (SPSS, Chicago, IL). The continuous data were expressed as means ± SD, and the correlations with the scintigraphy variable were tested by the Mann-Whitney nonparametric test. The various qualitative predictive information was crosstabulated with the scintigraphy variable, and the correlations were tested by the Pearson’s χ² test or the Fisher’s exact test. All P values were two-sided; a value <0.05 was considered statistically significant.

According to the results of exercise myocardial scintigraphy, 17 patients (21.2%) had SMI. None of these patients experienced angina pectoris during the test. Coronary angiography was performed in 14 of these patients (3 refused) and significant coronary stenoses were shown in 9 patients. SMI was significantly more frequent in men than in women (34.0 vs. 5.5%, P = 0.002) and in patients with neuropathy (32.6 vs. 17.6%, P = 0.012).

Two examples are shown in Fig. 1. Mean total counts of each level, and regional right-to-left ratios are shown in Table 2. A one-level abnormality was found in 29 patients, a two-level abnormality was found in 2 patients, and a three-level abnormality was found in 2 patients. The abnormalities were most commonly found on the calves (23 patients). The abnormalities were unilateral (14 on the right LL and 18 on the left LL), except in one patient, who had an abnormality of the right thigh as well as the left buttock. Therefore, LL ²⁰¹Tl scanning was abnormal in 33 patients.

The univariate correlates of the positive results of LL ²⁰¹Tl scanning were a high level of HbA_1c (P = 0.044) and the presence of retinopathy (P = 0.042) (Table 3). Of the 33 patients with an abnormal result of LL scintigraphy, 10 patients (30.3%) had abnormal results of myocardial scanning. Of these 10 patients, coronary angiography was performed in 8 patients; significant coronary stenoses were shown in 6 patients and normal coronary arteries were shown in 2 patients. Of the 47 patients with normal results of LL scanning, 7 patients (14.9%) had abnormal results of myocardial scanning. One patient refused the coronary angiography, three patients had significant coronary stenoses, and three patients had normal coronary arteries. An abnormal result of myocardial ²⁰¹Tl scintigraphy did not correlate significantly with an abnormal result of LL scintigraphy (P = 0.097).

The patients were separated into three groups. Group 1 consisted of the 47 patients with normal results of LL scanning. Group 2 consisted of the 13 patients with proximal (buttock and/or thigh levels) asymmetry of the LL perfusion but no abnormality at the calf level (these patients were considered not to have a macrovascular disease). Group 3 consisted of the 20 patients with isolated abnormal perfusion of the calf (Table 4). The prevalence of nephropathy differed significantly between the three groups (P = 0.016) and was significantly higher in group 2 (53.8%) than in group 1 (23.4%) (P = 0.032). There was a similar trend for retinopathy (P = 0.065), the prevalence of which was higher in group 2 (53.8%) than in group 1 (21.3%) (P = 0.023).

Arteriography is the gold standard for diagnosing occlusive arterial disease of the legs. It provides morphological data but no information about flow, which depends on the amount of exercise, the collateral circulation, and the presence of small-vessel disease. Furthermore, this test is invasive. ²⁰¹Tl is a potassium analog that has many of the physical characteristics of potassium. Its distribution within the skeletal muscle reflects the fractional distribution of cardiac output and is related to regional blood flow (24). Regional ²⁰¹Tl uptake in the LL is studied after exercise (11,12,13,14,15,16,19), as in myocardial perfusion studies. The reproducibility is high (15). Two methods are available: the first consists of scanning the whole body and comparing LL thallium activity with whole-body activity (11,12,14,15,16,19). Using this method, Segall et al. (11) compared patients with angiographically proved PVD and age-matched men without PVD. They showed that this method had a sensitivity of 100% (11). However, specificity was only 10%, and young men without PVD had to be used as control subjects to reach a specificity of 71% (11). To overcome that point, the second method has been used. It consists of measuring the regional ²⁰¹Tl uptake in the LL and calculating the right-to-left side intraindividual ratio (11,12,13,16,19). The physiological differences related to sex, age, weight, and physical training are then avoided (11). Tellier (12) studied a control population of 54 healthy male subjects with a mean age of 54 years and found an abnormal ratio after stress (defined by two SD over the mean value) for buttocks, thighs, and calves of 1.120, 1.078, and 1.084, respectively (12). The ratios were 1.04 and 1.2 for thighs and calves, respectively, when Seder et al. studied 11 healthy young men (16). A ratio <0.9 or >1.1 was considered by Segall et al. (11) to be abnormal. Indeed, such a ratio at one or several levels (buttocks, thighs, or calves) afforded a sensitivity of 81% for PVD detection as shown by angiography or a Doppler ankle-brachial index <1.0 (11). Using a control population of young men or age-matched men, the specificity was 83 and 80%, respectively (11).

In the present study, we demonstrated abnormal results of LL ²⁰¹Tl scanning in 42% of diabetic patients with a high cardiovascular risk profile. These abnormalities suggest peripheral vascular disorders at a silent stage, because they were found in asymptomatic patients who were able to reach the maximal theoretic heart rate during the stress testing. The prevalence of abnormal results of LL scintigraphy was twice the prevalence of SMI. This is in good agreement with the results of Tellier (12), who performed the same investigations in a population with proven coronary artery disease and in a group of smokers (19). As reported before (12,13,16,19), the lesions were essentially distal.

Abnormal results of LL ²⁰¹Tl scanning might have been expected to be associated with coronary stenoses in the patients with SMI. Such association was not found (P = 0.097), but further investigation is needed in a larger series. LL perfusion abnormalities may result not only from macrovascular disease but also from microvascular disorders. The fact that the occurrence of proximal perfusion defects without distal defects is associated with a higher prevalence of nephropathy and retinopathy is in agreement with microvascular disorders. Because we have previously found evidence for endothelial dysfunction and a reduction in coronary reserve due to microvascular changes in diabetic patients with SMI but angiographically normal coronary arteries (10), we may hypothesize that such factors might be involved in the patients with an abnormal result of LL scanning but without significant peripheral artery lesions. Therefore, femoral endothelial dysfunction has also been demonstrated in type 2 diabetic patients (25). Endothelial dysfunction might also contribute to the decrease in muscle ²⁰¹Tl uptake by interfering with distal vasomotricity. Although microvascular disease is likely to be bilateral, our findings suggest that, as in the eye, the disease may be asymmetrically distributed.

In conclusion, the present study using LL ²⁰¹Tl scintigraphy coupled with exercise in diabetic patients with other cardiovascular risk factors but no claudication shows evidence of defective peripheral blood flow in 42% of cases. Further study is needed to clarify whether this disorder is mainly related to macrovascular or microcirculatory dysfunction. Abnormal results of LL scanning correlated with poorer glycemic control and microangiopathic complications, which suggests that they might account for microvascular disease. The cost of LL imaging during myocardial scintigraphy is low because it requires only some additional time. Early scintigraphic detection of functional vascular abnormalities and their location and extent should lead to early and preventive management. Prospective studies are mandatory to evaluate the usefulness of early detection of PVD and its putative predictive value for more severe angiopathic lesions and complications.

We thank Sandrine Perret for secretarial assistance and Marion Sutton-Attali for assistance with the English manuscript.