Non-culprit plaque characteristics in acute coronary syndrome patients with raised hemoglobinA1c: an intravascular optical coherence tomography study

Study patients were retrospectively selected from the 2nd Affiliated Hospital of Harbin Medical University. Patients (aged ≥ 18 years) presenting with ACS and undergoing emergency procedures were screened for OCT examination. The main exclusion criteria were cardiogenic shock, end-stage renal disease, serious liver dysfunction, allergy to contrast media, and contraindication to aspirin or ticagrelor. Patients with left main disease, chronic total occlusion, or extremely tortuous or heavily calcified vessels were not included because of the potential difficulty in performing OCT in such situations. Clinical data including DM history, results of 75 g oral glucose tolerance test (OGTT) and HbA1c were recorded. In this study, 226 patients with ACS who underwent OCT imaging were included. Among these, 159 patients had serum HbA1c records, and 67 patients were diagnosed as DM according to OGTT (n = 10) or DM history (n = 57). Patients with poor image quality (n = 6) and left main disease (n = 4) were excluded. Finally, 216 patients were included in the analysis. The study flow chart is shown in Fig. 1. Non-culprit lesions were defined as plaques with angiographic diameter stenosis between 30 and 70% that had not been treated during the session of percutaneous coronary intervention (PCI) according to the results of stress test or ECG changes during the spontaneous ischemia attacks.

Patients were divided into three groups according to the serum level of HbA1c. Glucose metabolism was assessed according to ADA and World Health Organization (WHO) criteria [3].

Normal HbA1c (NA1c) was defined as plasma HbA1c value < 5.7%. Raised HbA1c (RA1c) was defined as plasma HbA1c value ≥ 5.7 but < 6.5% [3]. DM was defined as a history of DM, HbA1c value ≥ 6.5%, and fasting plasma glucose (FPG) level ≥ 126 mg/dL or 2-h plasma glucose (PG) level ≥ 200 mg/dL during OGTT [3]. All laboratory data were measured at fasting during the hospitalization. This study was approved by the Ethics Committee of the 2nd Affiliated Hospital of Harbin Medical University (Harbin, China), and all patients provided written informed consent.

Coronary angiography was performed after intra-coronary administration of 100–200 µg nitroglycerin. Quantitative coronary angiography (QCA) analysis was performed using Cardiovascular Angiography Analysis System (CAAS, 5.10, Pie Medical Imaging B.V., Maastricht, Netherlands). After selection of end-diastolic frames and calibration using the catheter’s tip, reference vessel diameter (RVD), minimal lumen diameter (MLD), diameter stenosis, and lesion length were measured.

A commercially available frequency-domain OCT system (OCT C7-XR Dragonfly, St. Jude Medical, St. Paul, MN) was used in the present study. With the help of a 6 or 7-F guiding catheter, an OCT imaging catheter was advanced distal to the lesion, and automated pullback was initiated in concordance with blood clearance by the injection of contrast media or low molecular dextran from the guiding catheter, and the OCT wire was pulled back at a rate of 20 mm/s. All images were de-identified and digitally stored.

Plaques were classified as fibrous (homogeneous, highly backscattering region) or lipid (low signal region with diffuse border). When lipid was present ≥ 90° in any of the cross-sectional images within the plaque, it was considered a lipid-rich plaque. The lipid arc was measured at every 1-mm interval throughout the length of each lesion, and the values were averaged. Lipid length was also measured on longitudinal view. Lipid index (LI) was defined as the mean lipid arc multiplied by lipid length. The fibrous cap thickness (FCT) of a lipid-rich plaque was measured 3 times at its thinnest part and the average value was calculated. A TCFA was defined as the thinnest fibrous cap with a thickness ≤ 65 μm in a lipid-rich plaque on cross-sectional imaging [6]. Each plaque was separated by at least 5 mm from the edge of another plaque or an implanted stent edge.

Calcification was an area, which consisted of a signal-poor or heterogeneous region with a sharply delineated border [7]. The number and longitudinal length of individual calcium deposits were recorded. Quantitative analysis was performed using cross-sectional OCT images at 1-mm intervals. Calcium deposits were analyzed individually by measuring calcium arc and depth (minimum distance from lumen to superficial calcium edge). Calcium index was calculated for each calcium deposit as the product of calcium length and mean calcium arc. Total calcium index was calculated as the sum of the calcium index of each calcium deposit within the whole lesion segment. The mean calcium index was calculated as the total calcium index divided by the number of calcium deposits. Spotty calcium deposits were defined as those with length < 4 mm and maximal arc < 90°, and deposits not meeting these criteria were classified as large calcium deposits.

Macrophage infiltration was defined as signal-rich, distinct or confluent punctuate regions that exceed the intensity of background spackle noise [7]. Microchannels were defined as signal-poor voids that were sharply delineated in multiple contiguous frames [8]. Plaque disruption was identified by the presence of fibrous cap discontinuity with a clear cavity formation inside the plaque.

The OCT data were analyzed by two experienced investigators who were blinded to the angiographic and clinical findings. When there was a difference between the investigators, a consensus reading was obtained from a third independent reviewer. The inter- and intra-observer kappa coefficients for lipid arc were 0.896 and 0.901, respectively.

Baseline characteristics are presented in Table 1. The prevalence of hypertension was significantly higher in the DM group than in the other two groups (P = 0.046). The frequency of dyslipidemia was significantly higher in the RA1c and DM groups than that in the NA1c group (P = 0.021). The creatinine level and the prevalence of chronic kidney disease (CKD) were comparable among the three groups. There was no difference in statin treatment among the three groups. In the DM group, 7 subjects (8.9%) were treated with insulin therapy, and 11 subjects (13.9%) were treated with oral agents.

Angiographic findings are shown in Table 2. A total of 305 non-culprit plaques were detected in 216 subjects (95 plaques in 68 NA1c subjects, 93 plaques in 69 RA1c subjects, and 117 plaques in 79 DM subjects). The plaque distribution was not significantly different among the three groups (P = 0.966). Plaques in patients with RA1c and DM had a longer lesion length (P = 0.008). Patients with RA1c had a larger stenosis diameter percentage than the other two groups (P = 0.020), but the difference was slight.

Representative OCT images in each group are shown in Fig. 2. A comparison of the quantitative OCT findings of the lipid plaques between the NA1c, RA1c, and DM groups is presented in Fig. 3. The plaques had a significantly longer lipid length and greater LI in the RA1c and DM groups than in the NA1c group (lipid length: 17.0 ± 8.3 mm [RA1c] vs. 13.9 ± 7.2 mm [NA1c], P = 0.004 and 16.3 ± 7.8 mm [DM] vs. 13.9 ± 7.2 mm [NA1c], P = 0.009; LI: 2755.0 ± 1694.0 mm° [RA1c] vs. 1592.1 ± 981.2 mm° [NA1c], P = 0.001 and 2758.0 ± 1821.7 mm° [DM] vs. 1592.1 ± 981.2 mm° [NA1c], P = 0.001). Compared to patients with NA1c, plaques in those with DM had a significantly wider maximum lipid arc and mean lipid arc (maximum lipid arc: 246.8 ± 90.6° [DM] vs. 199.7 ± 79.9° [NA1c], P = 0.012; mean lipid arc: 175.2 ± 63.7° [DM] vs. 142.2 ± 57.1° [NA1c], P = 0.012). Additionally, the maximum and mean lipid arcs were wider in patients with RA1c than in those with NA1c with a borderline significance (maximum lipid arc: 232.4 ± 91.3° [RA1c] vs. 199.7 ± 79.9° [NA1c], P = 0.090; mean lipid arc: 164.7 ± 63.7° [RA1c] vs. 142.2 ± 57.1° [NA1c], P = 0.090). There were no significant differences in these lipid plaque parameters between the RA1c and DM groups.

The FCT of plaques in RA1c subjects and DM subjects tended to be thinner than that in NA1c subjects, although there was no significant difference (107.2 ± 61.9 mm [NA1c] vs. 87.0 ± 44.2 mm [RA1c] vs. 91.1 ± 47.9 mm [DM], P = 0.167).

In univariable linear regression (Table 3), HbA1c was significantly correlated with maximal lipid arc (coefficient β = 15.469, P = 0.002) and lipid length (coefficient β = 1.678, P < 0.001). All the variables with P < 0.1 were entered en bloc in the multivariable model. As result, the model of maximal lipid arc was adjusted by age, smoking, HbA1c, triglyceride (TG) and creatinine, and the model of lipid length was adjusted by HbA1c, TG and high-density lipoprotein cholesterol (HDL-C). In the multivariable linear regression analysis (Table 4), a higher HbA1c was independently associated with a larger maximal lipid arc (coefficient β = 14.282, P = 0.005) and a longer lipid length (coefficient β = 1.368, P = 0.003).

The prevalence of OCT plaque characteristics are shown in Table 5. Of the 305 non-culprit plaques, 157 were found to contain a lipid core (40, 43 and 54 in the NA1c, RA1c and DM groups, respectively). The prevalence of lipid-rich plaques was similar among NA1c, RA1c and DM groups (P = 0.735).

The frequency of calcification in RA1c group (P = 0.048) and DM group (P = 0.001) was higher than that in NA1c group. There was a significant difference in the prevalence of spotty calcification between DM group and NA1c group (36.8% vs. 20.0%, P = 0.006). Detailed OCT analysis of calcium deposits is summarized in Table 6.

The prevalence of TCFA (P = 0.335), disruption (P = 0.689), thrombus (P = 0.696) and microchannels (P = 0.419) was not different among the three groups. The prevalence of macrophage infiltration in the DM group was higher than that in the NA1c group (P = 0.005). Although the frequency of macrophage infiltration tended to be higher in RA1c subjects than in NA1c subjects, the difference was not significant (P = 0.215).

In the present study, there were several limitations. First, this was a retrospective study using a registry database. Therefore, potential selection bias is unavoidable. Patients with cardiogenic shock (hemodynamic instability), congestive heart failure, chronic total occlusion, left main disease, or renal failure were not included. Second, FPG and 2-h PG concentrations may be increased in ACS, which increases the risk of inadvertently classifying patients with non-DM as having DM. Third, the exact measurements of necrotic core and plaque burden by OCT were not possible because of the relatively shallow axial penetration. However, because the most important morphological determinants of plaque vulnerability are superficial, the region of greatest interest was still within the imaging range of current OCT systems. Finally, the impairment of duration of RA1c to plaque vulnerability was not considered in this study.