2015 Volume 79 Issue 4 Pages 880-888
Background: Aberrant expression of microRNAs is associated with neointimal hyperplasia (NIH) in type 2 diabetes. We prospectively compared the effects of pioglitazone on coronary NIH and changes in microRNAs according to NIH status in type 2 diabetic patients during 9-month follow-up.
Methods and Results: Type 2 diabetic patients were randomly assigned to the pioglitazone (n=36) or control groups (n=36) after coronary stenting. Primary endpoint was the comparison of changes in neointimal volume on OCT and in the level of circulating microRNA-17,-24,-92a,-126 and -145 during 9-month follow-up. Secondary endpoint was the comparison of changes in brachial artery flow-mediated dilation and inflammatory markers such as IL-6, TNF-α, hsCRP, adiponectin, sICAM-1, and sVCAM-1 between the 2 groups. Neointimal volume was significantly lower in the pioglitazone group (25.02±17.78 mm3 vs. 55.10±30.01 mm3, P<0.001) with significant increases in circulating microRNA-24 (0.264±0.084 vs. 0.006±0.030, P<0.001) during follow-up. FMD was significantly greater in the pioglitazone than control group at 9 months (0.47±0.14 mm vs. 0.28±0.18 mm, P<0.05, respectively). Decreases in inflammatory markers such as IL-6, TNF-α, and sVCAM-1 were significantly greater in the pioglitazone than the control group during the follow-up.
Conclusions: Pioglitazone significantly decreased NIH with increases in circulating microRNA-24 at 9-month follow-up. The decrease in microRNA-24 could be used as a potential predictor of increases in NIH in type 2 diabetic patients. (Circ J 2015; 79: 880–888)
Type 2 diabetes with systemic inflammation aggravates endothelial dysfunction and coronary artery disease, leading to cardiovascular events.1–3 Type 2 diabetes with coronary artery disease requiring stent implantation is associated with a higher chance of in-stent restenosis.4,5 Endothelial dysfunction in type 2 diabetes is associated with the progression of atherosclerosis and neointimal proliferation after coronary stenting.6–8 A diabetes medication that also improves endothelial dysfunction and reduces neointimal proliferation would be of significant clinical benefit to diabetic patients after coronary stent implantation. Increasing interest remains in the identification of systemic pharmacological therapies to improve endothelial dysfunction and prevent neointimal hyperplasia after coronary stenting. In our previous study, early decreases in the number of circulating inflammatory cells together with circulating tumor necrosis factor (TNF)-α, interleukin (IL)-6, and MCP-1 concentration have been noted after pioglitazone treatment.9 The early decreases in smooth muscle cell migration and proliferation with pioglitazone treatment have been observed in type 2 diabetic patients.9
Editorial p 768
Optical coherence tomography (OCT) with significantly higher resolution than intravascular ultrasound has not been used in the evaluation of neointimal characteristics in type 2 diabetic patients after pioglitazone treatment. Moreover, the effects of pioglitazone in brachial artery endothelial function and circulating levels of microRNAs associated with endothelial function and smooth muscle cell migration have not been evaluated. MicroRNAs are post-transcriptional regulators that bind to complementary sequences on target mRNAs.10,11 Aberrant expression of microRNAs is implicated in numerous disease states. MicroRNA-17,-24,-92a,-126,-145 are expressed in endothelial and vascular smooth muscle cells with critical roles in endothelial function.12–18 Expression of endothelial microRNA-24 reduces the organization of actin into stress fiber, thereby decreasing cell migration;18 furthermore, stable plaque expresses higher microRNA-24 than unstable plaque.18,19 We prospectively compared the effects of pioglitazone in neointimal hyperplasia after coronary stenting and the changes in circulating microRNA-17,-24,-92a,-126,-145 with neointimal hyperplasia status in type 2 diabetic patients during 9-month follow-up.
Patients were eligible for this study if they were 45–75 years of age and had both type 2 diabetes and significant coronary artery disease requiring stent implantation. Both previously treated and untreated type 2 diabetic patients after coronary stenting were prospectively included in this study at Korea University Anam Hospital cardiovascular centers from October 2011 through February 2013. A total of 239 patients were screened for inclusion in the study. Patients (n=55) who did not fulfill the inclusion criteria or who had any of the exclusion criteria (n=112) were excluded (Figure 1). Eligible patients (n=72; 32 women and 40 men) were randomly assigned to receive either pioglitazone 15 mg (n=36) or control (n=36) after coronary stenting. To be included in the study, fasting plasma glucose ≥126 mg/dl or random plasma glucose ≥200 mg/dl in a patient with classic symptoms of hyperglycemia or hyperglycemic crisis had to be documented for previously untreated type 2 diabetic patients. We excluded patients with the following conditions: use of pioglitazone within 3 months before enrollment, severe hypertension (systolic blood pressure >180 mmHg and diastolic blood pressure >110 mmHg), acute myocardial infarction, severe left main coronary disease, unsuccessful reperfusion after stenting, and previous history of coronary stenting or bypass surgery. Patients with heart failure (ejection fraction <45% or signs of heart failure), hepatic dysfunction (serum aspartate or alanine aminotransferase above twice the upper limit of normal ranges), serum creatinine >2.0 mg/dl, and life expectancy <1 year were also excluded.
Study protocol. AMI, acute myocardial infarction; baFMD, brachial artery flow-mediated dilation; CABG, coronary artery bypass grafting; f/u, follow-up; OCT, optical coherence tomography; PCI, percutaneous coronary intervention; PPAR-γ, peroxisome proliferator-activated receptor gamma.
The primary endpoints of this study were comparison of changes in neointimal volume on OCT and in the levels of circulating microRNA-17,-24,-92a,-126 and -145, which are known as indicators of endothelial cell migration and atherosclerosis progression, during 9-month follow-up. Secondary endpoints were comparison of changes in brachial artery flow-mediated dilation (baFMD) between the 2 groups during 9-month follow-up. Inflammatory markers such as high-sensitivity C-reactive protein (hsCRP), IL-6, TNF-α, adiponectin, soluble intercellular adhesion molecule-1 (sICAM-1), and soluble vascular cell adhesion molecule-1 (sVCAM-1), as well as the insulin resistance index such as the homeostasis model of assessment (HOMA) index and retinol-binding protein-4 (RBP4), were compared during the 9-month follow-up. Body mass index was calculated by dividing the square of the height in meters by the weight in kilograms. The study was approved by the University Hospital Institute Review Board, and all participants gave written informed consent.
Coronary Angiography and OCTAll patients were asked to return after 9 months for angiographic follow-up. If clinically indicated, follow-up angiography was performed earlier. All participating patients received everolimus-eluting stents (Xience Prime®; Abbott Lab., IL, USA) according to the study protocol. Procedural success was defined as residual stenosis <15% in the absence of closure during the first 48 h after the procedure. Coronary angiograms were obtained at baseline, immediately after stenting, and at 9-month follow-up. Two identical orthogonal views were obtained after i.c. nitrate treatment and stored on digital CD-ROM. All angiographic and clinical data were analyzed by individuals who were unaware of the treatment assignments. End-diastolic frames were chosen for quantitative analysis, which was performed using a computer-based TCS system, Version 2.02 (Medcon, Tel-Aviv, Israel). The reference diameter, minimal luminal diameter, percentage of stenosis, and lesion length were calculated as the average of the 2 orthogonal views. The same views and calibrations were used at follow-up angiography. The average diameter of normal segments proximal and distal to the treated lesion was used as the reference diameter. Restenosis was defined as stenosis >50% of the luminal diameter. Balloon angioplasty and stent implantation were performed according to standard clinical practice, as described previously.20
OCT data were analyzed at the Korea University OCT Core Laboratory, and all OCT data were analyzed by individuals who were unaware of the treatment assignments. OCT was performed after 200-μg i.c. nitroglycerin injection. OCT images were acquired using a non-occlusive technique with the C7XR system (LightLab Imaging, Westford, MA, USA), and images were digitally stored for subsequent analysis. Mean area and volumes of lumen, stent, and neointimal hyperplasia were calculated along the entire stented segment. Quantitative strut-level OCT analysis was performed at 0.33-mm intervals. The center of the luminal surface of the strut was determined for each strut, and its distance to the lumen contour was calculated to determine strut-level neointimal thickness. The number of struts without coverage was counted for each frame in order to count the total number of uncovered struts per lesion. Struts were categorized as uncovered when a tissue layer on the endoluminal surface was not visible; as covered embedded struts when covered by tissue and not interrupting the smooth lumen contour; as covered rhombus struts when covered by tissue but extending into the lumen; and as malapposed if the distance from the endoluminal surface of the strut to the adjacent lumen contour was greater than the sum of the metal and polymer thickness, which was 90 μm for the everolimus-eluting stent. Neointima was the tissue between the luminal border and the inner border of the struts.
For qualitative OCT analysis, the peri-strut low-intensity area, which was defined as an area around stent struts that was homogenously hypointense to the surrounding tissue without significant signal attenuation behind the area, neovascularization and the presence of thrombus were evaluated on follow-up OCT.3 Neovascularization was defined as a small black hole or a tubular structure within a plaque. For the measurement of neointimal unevenness score in each cross-section, the maximal neointimal thickness in 1 cross-section was divided by the average neointimal thickness of the cross-section.3
MicroRNA-17,-24,-92a,-126,-145Peripheral blood samples (5 ml) were drawn into serum collection tubes, were allowed to stand for approximately 30 min at room temperature and were centrifuged at 1,800 g for 10 min at room temperature. The serum was aliquoted into eppendorf tubes and stored at –80℃. Total RNAs from human serum were isolated using TRI Reagent BD (MRC, TB126) following the instructions from the manufacturer with modification. In brief, 250 μl of serum per eppendorf tube was added to 0.75 ml of TRI Reagent BD, and was stored for 5 min at room temperature. The samples were extracted with 200μl chloroform, and the aqueous phase containing RNA was transferred to a fresh tube. The RNA was precipitated from the aqueous phase by centrifugation at 12,000 g for 15 min at 4℃ after mixing with 500 μl of isopropanol. The RNA pellet was washed in 1 ml of 75% ethanol by centrifugation, and finally the pellet was re-suspended in 5 μl of RNase-free water. The samples isolated from the same patients were gathered. Total RNA was quantified using spectrophotometry (ND-1000; NanoDrop Technologies, Wilmington, DE, USA). Ten ng of total RNA isolated from the serum was reverse transcribed in a 15-μl reaction volume using the TaqMan microRNA Reverse Transcription kit (Applied Biosystems, Grand Island, NY, USA) according to the manufacturer instructions. MiRNA-specific stem-loop reverse transcription (RT) primers, miR-17, miR-24, miR-92a, miR-126, miR-145 and miR-16 primers were used for RT reaction. Subsequently, 2 μl of the RT product was used for detecting miRNA expression on quantitative polymerase chain reaction (qPCR) using TaqMan microRNA Assay kits (Applied Biosystems) for the corresponding microRNA. Real-time PCR was performed using iQTM Cycler (Bio-Rad Laboratories, CA, USA) using the following program: 10 min pre-incubation at 95℃ and 40 cycles of 15 s of denaturation at 95℃ and 60 s of annealing/extending at 60℃. MiR-17, miR-24, miR-92a, miR-126 and miR-145 primers and miR-16 primers as an endogenous control were used. The amount of miRNA not detected after 40 cycles of real-time PCR was regarded in the present study as a CT equivalent to 40. Negative controls were included with every real-time RT-PCR assay, and no amplification of the signal was observed when water was added instead of RNA or cDNA. The measurement of miRNA expression was assayed in duplicate. Ct was normalized to miR-16 and expressed as 2–(Ct[microRNA]–Ct[miR-16]).
baFMDPatients fasted for at least 8 h before baFMD measurement; moreover, patients were educated not to exercise and not to ingest substances that might affect baFMD such as caffeine, high-fat foods and vitamin C or use tobacco for at least 4 h before the study. Echocardiography (Vivid q; GE Healthcare, WI, USA) was used for acquiring images, and a linear array transducer (12L; GE Healthcare) was used to acquire images with sufficient resolution. Patients were positioned supine with the arm in a comfortable position for imaging the brachial artery. The brachial artery was imaged above the antecubital fossa in the longitudinal plane. A segment with clear anterior and posterior intimal interfaces between the lumen and vessel wall was selected for continuous 2-D grayscale imaging, and brachial artery diameter was measured at end diastole with the electrocardiogram r wave as a marker for 3 cardiac cycles. All measurements were averaged for analysis. During image acquisition, anatomic landmarks such as veins and fascial planes were noted to help maintain the same image of the artery during 9-month follow-up. To create a flow stimulus in the brachial artery, a blood pressure cuff was first placed above the forearm. Baseline rest images were acquired, and blood flow was estimated by time-averaging the pulsed Doppler velocity signal obtained from a midartery sample volume. Thereafter, arterial occlusion was created by cuff inflation to 220 mmHg above systolic pressure to occlude arterial inflow for 5 min. Subsequent cuff deflation induced a brief high-flow state through the brachial artery, resulting in increase in shear stress with the brachial artery dilation. The longitudinal image of the artery was recorded continuously from 30 s before to 2 min after cuff deflation to measure percent change in diameter. A midartery pulsed Doppler signal was obtained upon immediate cuff release and at 10 s after cuff deflation to assess hyperemic velocity. After baseline conditions were reestablished 15 min after endothelium-dependent vasodilation, 0.6 mg nitroglycerin was administered under the tongue for the measurements of endothelium-independent vasodilation. Intraobserver correlation for brachial artery dilation for 10 random cases was 0.94.
Inflammatory Markers, Insulin Resistance and Lipid ProfileVenous blood samples were drawn from each patient after 8 h or overnight fasting. Blood samples were centrifuged to obtain plasma, and the plasma was stored at –80℃. Plasma glucose was measured using the glucose oxidase method, and serum insulin was measured using immunoradiometric assay (Biosource, Nivelles, Belgium). The HOMA index, a parameter of insulin resistance, was calculated using baseline glucose and insulin as follows: fasting glucose (mmol/L)×fasting insulin (μU/ml)/22.5. Serum RBP4 was assessed using enzyme-linked immunosorbent assay (ELISA) from a commercial kit (AdipoGen, Seoul, Korea). In this assay, which is based on an RBP4 competitive ELISA system, 100 μg/L recombinant human RBP4 expressed by an animal cell line (HEK293 cells), was coated onto a plate, and varying levels of the recombinant RBP4 and a polyclonal antibody were added to the plate and were used to produce a standard curve. The intra- and inter-assay coefficients of variation for this ELISA were 6.0% and 7.9%, respectively.
Inflammatory markers such as hsCRP, IL-6, TNF-α, adiponectin, sICAM-1, and sVCAM-1 were measured in both groups at the beginning of the study and at 9-month follow-up. TNF-α was measured on sandwich ELISA with a minimum detectable level of 0.5 pg/ml (ALPCO Diagnostics, Salem, NH, USA). Undetectable TNF-α was recorded as 0.4 pg/ml for 2 patients. IL-6 was also measured on sandwich ELISA with a minimum detectable level of 0.16 pg/ml (ALPCO Diagnostics). hsCRP concentration was quantified using a latex nephelometer II (Dade Behring, Newark, DE, USA). Plasma adiponectin concentration was assessed on radioimmunoassay (Linco Research, St. Charles, MO, USA). The sensitivity of this assay was 0.78 ng/ml. The coefficients of variation for intra- and inter-assay were 9.3% and 15.3%, respectively. In addition, sICAM-1 and sVCAM-1 were measured using a commercially available ELISA according to the manufacturer’s instructions (R & D Systems, Minneapolis, MN, USA). Total cholesterol, triglyceride, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol (LDL-C) were measured using enzymatic methods with standard biochemical procedures on a B.M. Hitachi automated clinical chemistry analyzer (Hitachi, Tokyo, Japan).
Statistical AnalysisData are expressed as mean±SD for continuous variables, and as number and percentage for the categorical variables. Fisher’s exact test or a chi-squared test was used for categoric variables. The change from baseline was calculated as that obtained at the end of treatment subtracted from the value obtained at the beginning of the intervention. The results between 2 groups were compared using unpaired Student’s t-test, and the comparisons between before and after treatment were analyzed on paired t-test. OCT and angiographic analyses were performed according to the number of patients available for each analysis. The sample size of the study was determined based on estimation of the primary endpoint of neointimal volume from our previous trial:20 in the pioglitazone group, we assumed neointimal volume of 1.9±1.0 mm3/1 mm, and in the control group, we assumed neointimal volume of 2.6±1.4 mm3/1 mm. Using a 2-sided test for differences in independent binomial proportions with an α level of 0.05, we calculated that 58 patients (29 patients for each group) would be required for the study to have 80% power to detect a difference in neointimal volume between 2 groups; therefore, we enrolled 36 patients in each group to account for 20% loss in the OCT follow-up. Variables that did not show a normal distribution were log-transformed for subsequent analyses. The results were analyzed according to the intention-to-treat principle, and the initial pioglitazone treatment was used for analysis. P<0.05 was considered significant. SPSS (version 20.0) was used for analyses (IBM SPSS, New York, NY, USA).
Baseline patient characteristics of the pioglitazone group (n=36) and the control group (n=36) were similar (Table 1). Mean age was similar (58.3±11.8 and 60.0±12.7 years) as was mean body mass index (24.6±3.9 and 24.4±3.5 kg/m2) between the pioglitazone and control groups. The rates of hypertension, hyperlipidemia, current smoking, and family history of coronary artery disease at baseline did not show significant differences between the 2 groups. The number of patients taking oral hypoglycemic medication (80.6%, n=29 in the pioglitazone group vs. 86.1%, n=31 in the control group, P=0.527) was similar between the 2 groups. The rates of various active medications at baseline did not show significant differences. Pioglitazone at baseline was doubled during follow-up in 7 patients (19.4%) in the pioglitazone group at physician discretion. No incidence of fracture or bladder cancer was noted during the 9-month follow-up in both groups (Table S1).
| Characteristics | Pioglitazone group (n=36) |
Control group (n=36) |
P-value |
|---|---|---|---|
| Age (years) | 58.3±11.8 | 60.0±12.7 | 0.282 |
| Male sex | 19 (52.8) | 21 (58.3) | 0.635 |
| BMI (kg/m2) | 24.6±3.9 | 24.4±3.5 | 0.801 |
| Risk factors | |||
| Hypertension | 14 (38.9) | 13 (36.1) | 0.808 |
| Hyperlipidemia | 12 (33.3) | 13 (36.1) | 0.804 |
| Current smoking | 9 (25.0) | 7 (19.4) | 0.571 |
| Family history of CAD | 5 (13.9) | 8 (22.2) | 0.358 |
| Past history of TIA or stroke | 1 (2.8) | 1 (2.8) | 1.000 |
| LVEF (%) | 57.1±9.5 | 56.9±10.1 | 0.894 |
| Stable angina | 20 (55.6) | 18 (50.0) | 0.637 |
| Unstable angina | 16 (44.4) | 18 (50.0) | 0.637 |
| Duration of diabetes (months) | 28±24 | 26±24 | 0.676 |
| Medication at baseline | |||
| Oral anti-diabetic therapy | 29 (80.6) | 31 (86.1) | 0.527 |
| Biguanides | 20 (55.6) | 24 (66.7) | 0.334 |
| α-Glucosidase inhibitors | 8 (22.2) | 11 (30.6) | 0.422 |
| Sulfonylureas | 16 (44.4) | 17 (47.2) | 0.813 |
| Insulin | 5 (13.9) | 4 (11.1) | 1.000 |
| Aspirin | 26 (72.2) | 22 (61.1) | 0.317 |
| ACEI | 3 (8.3) | 4 (11.1) | 1.000 |
| ARG | 11 (30.6) | 9 (25.0) | 0.599 |
| β-blocker | 9 (25.0) | 6 (16.7) | 0.384 |
| Calcium channel blocker | 12 (33.3) | 14 (38.9) | 0.624 |
| Diuretics | 5 (13.9) | 3 (8.3) | 0.710 |
| Nitrate | 13 (36.1) | 17 (47.2) | 0.339 |
| Nicorandil | 3 (8.3) | 3 (8.3) | 1.000 |
Data given as n (%) or mean±SD. ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; CAD, coronary artery disease; LVEF, left ventricular ejection fraction; TIA, transient ischemic attack.
Quantitative OCT and angiographic analyses are given in Tables 2,S2. Neointimal volume, which was the primary endpoint, was lower for pioglitazone than control (25.02±17.78 mm3 vs. 55.10±30.01 mm3, P<0.001). The rates of uncovered well-apposed struts (3.2±4.0% vs. 2.9±4.8%, P=0.837) and uncovered malapposed struts (0.1±1.3% vs. 0.1±1.1%, P=0.914) per lesion were similar between the 2 groups. Similarly, the rates of covered embedded struts (94.9±8.1% vs. 95.1±8.7%, P=0.879) and covered rhombus struts (1.7±2.8% vs. 2.0±3.2%, P=0.803) per lesion were similar between the 2 groups. Mean neointimal thickness of covered struts was significantly lower for pioglitazone than control (0.16±0.15 mm vs. 0.28±0.34 mm, P<0.001). Significant negative correlations between neointimal volume and microRNA-24, FMD at 9 months, and plasma adiponectin concentration are shown in Figure S1.
| Variable | Pioglitazone (n=36) |
Control (n=36) |
P-value |
|---|---|---|---|
| No. patients with 9-month follow-up | 28 (77.8) | 26 (72.2) | 0.586 |
| No. target lesions | 38 | 38 | |
| Mean stent length (mm) | 26.3±6.8 | 27.7±5.8 | 0.155 |
| Neovascularization | 2 (7.1) | 3 (11.5) | 0.663 |
| Frequency of intracoronary thrombus | 2 (7.1) | 1 (3.8) | 1.000 |
| Cross-section level analysis | |||
| No. struts analyzed per cross section | 6.7±1.9 | 6.5±1.8 | 0.872 |
| Mean lumen area (mm2) | 5.85±2.07 | 5.08±1.88 | <0.001 |
| Mean stent area (mm2) | 6.78±2.34 | 6.98±2.19 | 0.768 |
| Mean neointimal area (mm2) | 0.93±0.78 | 1.90±1.43 | <0.001 |
| Lumen volume (mm3) | 157.23±79.44 | 143.61±67.04 | 0.021 |
| Stent volume (mm3) | 181.43±104.91 | 196.32±110.19 | 0.115 |
| Neointimal volume (mm3) | 25.02±17.78 | 55.10±30.01 | <0.001 |
| Percentage net volume obstruction (%) | 13.9±10.1 | 28.5±13.4 | <0.001 |
| Strut-level analysis | |||
| Total no. analyzed struts (total) | 15,820 | 16,654 | |
| No. covered struts (total) | 15,283 | 16,203 | |
| Covered struts per lesion (%) | 96.7±5.3 | 97.0±6.2 | 0.780 |
| Covered embedded struts | 94.9±8.1 | 95.1±8.7 | 0.879 |
| Covered rhombus struts | 1.7±2.8 | 2.0±3.2 | 0.803 |
| No. uncovered struts (total) | 537 | 451 | |
| Uncovered struts per lesion (%) | 3.3±5.2 | 3.0±5.8 | 0.803 |
| Uncovered well-apposed struts | 3.2±4.0 | 2.9±4.8 | 0.837 |
| Uncovered malapposed struts | 0.1±1.3 | 0.1±1.1 | 0.914 |
| Mean neointimal thickness of covered struts (mm) | 0.16±0.15 | 0.28±0.34 | <0.001 |
| Neointimal unevenness score | 1.68±0.31 | 1.73±0.35 | 0.571 |
| Peri-strut low-intensity area (%) | 2.92±1.75 | 3.12±1.65 | 0.666 |
Data given as n (%) or mean±SD. OCT, optical coherence tomography.
MicroRNA-24, unlike microRNA-17, -92a, -126, and -145, increased significantly in the pioglitazone group when compared with the control group during 9-month follow-up (0.264±0.084 vs. 0.006±0.030, P<0.001; Figure 2A). On receiver operating characteristic curve analysis for the changes in microRNA-24 in detecting neointimal volume >25 mm3, the cut-off was 0.1715 with sensitivity of 0.861 and specificity of 0.944 (Figure 2B).
Change in microRNA during follow-up. (A) MicroRNA-24 increased significantly only in the pioglitazone group at 9-month follow-up. No significant differences were found for changes in microRNA-17,-92a,-126,-145. (B) Receiver operating characteristic curve and the corresponding area under the curve for change in microRNA-24 in detecting neointimal volume >25 mm3. Cut-off for microRNA-24 change was 0.1715 with a sensitivity of 0.861 and specificity of 0.944.
Increase in baFMD during 9-month follow-up was significantly greater for pioglitazone than control (0.47±0.14 mm vs. 0.28±0.18 mm, P<0.05, respectively; Table 3). Nitroglycerin-mediated dilation during the 9-month follow-up showed no significant differences between the pioglitazone and control groups (0.58±0.23 mm vs. 0.60±0.24 mm, P=0.804, respectively).
| Variables | Pioglitazone group (n=36) | Control group (n=36) | ||
|---|---|---|---|---|
| Baseline | At 9 months | Baseline | At 9 months | |
| Brachial artery diameter at rest (mm) | 3.96±0.42 | 4.00±0.40 | 3.99±0.39 | 4.01±0.41 |
| FMD (mm) | 4.18±0.41 | 4.46±0.39*,# | 4.25±0.38 | 4.30±0.40 |
| Change from at rest (mm) | 0.22±0.11 | 0.47±0.14*,# | 0.25±0.17 | 0.28±0.18 |
| Nitroglycerin-mediated dilation (mm) | 4.48±0.48 | 4.58±0.45 | 4.52±0.47 | 4.60±0.46 |
| Change from at rest (mm) | 0.53±0.20 | 0.58±0.23 | 0.53±0.22 | 0.60±0.24 |
Data given as mean±SD. *P<0.05 compared with baseline; #P<0.05 compared with the control group. baFMD, brachial artery flow-mediated dilation.
Decreases in inflammatory markers such as IL-6 (–2.57±2.19 pg/ml vs. –1.87±1.71 pg/ml, P<0.05, respectively), TNF-α (–4.02±1.77 pg/ml vs. –1.52±1.37 pg/ml, P<0.05, respectively), and sVCAM-1 (–207±213 ng/ml vs. 2±460 ng/ml, P<0.05, respectively) were significantly greater in the pioglitazone group when compared with the control group at 9-month follow-up (Table 4). CRP decreased significantly in both groups at 9-month follow-up compared with baseline, but the change from baseline was not significantly different between the 2 groups. Moreover, significant increases in adiponectin concentration was found in the pioglitazone group when compared with the control group (4.01±2.93 μg/ml vs. 0.23±1.15 μg/ml, P<0.05, respectively). RBP4, HOMA index and hemoglobin A1c (HbA1c) significantly decreased in both groups at 9-month follow-up, while the changes from baseline were similar.
| Variables | Pioglitazone group (n=36) | Control group (n=36) | ||
|---|---|---|---|---|
| Baseline | At 9 months | Baseline | At 9 months | |
| IL-6 (pg/ml)† | 4.37±4.01 | 1.81±1.31#,* | 4.77±3.94 | 2.90±2.28* |
| Change from baseline (pg/ml) | −2.57±2.19# | −1.87±1.71 | ||
| TNF-α (pg/ml)† | 6.83±4.76 | 2.82±3.05#,* | 6.16±5.27 | 4.61±3.60* |
| Change from baseline (pg/ml) | −4.02±1.77# | −1.52±1.37 | ||
| hsCRP (mg/L)† | 4.18±3.01 | 1.24±1.22* | 4.56±4.10 | 1.52±1.60* |
| Change from baseline (mg/L) | −2.93±2.62 | −3.03±3.09 | ||
| Adiponectin (μg/ml)† | 3.98±3.99# | 7.98±5.65#,* | 5.41±4.66 | 5.65±5.32 |
| Change from baseline (μg/ml) | 4.01±2.93# | 0.23±1.15 | ||
| sICAM-1 (ng/ml)† | 742±501 | 657±508 | 575±432 | 502±337 |
| Change from baseline (ng/ml) | −85±80 | −75±94 | ||
| sVCAM-1 (ng/ml)† | 976±588 | 769±393#,* | 1,065±692 | 1,069±811 |
| Change from baseline (ng/ml) | −207±213# | 2±460 | ||
| Fasting insulin (μU/ml)† | 12.8±4.5 | 9.4±3.8* | 13.2±6.0 | 10.0±6.4* |
| Change from baseline (pmol/L) | −3.4±3.5 | −3.1±3.3 | ||
| Fasting glucose (mmol/L)† | 7.9±3.1 | 6.3±1.9* | 8.0±3.2 | 6.4±2.0* |
| Change from baseline (mmol/L) | −1.6±2.3 | −1.5±3.1 | ||
| HOMA index† | 4.5±4.5 | 2.6±2.3* | 4.7±4.3 | 2.8±2.5* |
| Change from baseline (%) | −1.9±2.2 | −1.9±2.0 | ||
| HbA1c (%)† | 7.4±1.6 | 6.8±0.9* | 7.5±1.9 | 6.9±0.8* |
| Change from baseline (%) | −0.6±0.9 | −0.6±0.7 | ||
| RBP4 (μg/ml)† | 70.2±20.2 | 54.5±21.1* | 67.9±22.8 | 49.9±19.6* |
| Change from baseline (μg/ml) | −15.9±5.7 | −17.8±6.0 | ||
| TC (mg/dl) | 214±60 | 161±47* | 219±48 | 156±42* |
| Change from baseline (mg/dl) | −53±60 | −63±56 | ||
| LDL-C (mg/dl) | 149±66 | 90±46* | 159±77 | 89±45* |
| Change from baseline (mg/dl) | −60±45 | −68±55 | ||
| HDL-C (mg/dl) | 39±29 | 43±22 | 37±28 | 40±19 |
| Change from baseline (mg/dl) | 3±10 | 3±8 | ||
| Triglyceride (mg/dl)† | 135±99 | 119±79 | 129±83 | 123±60 |
| Change from baseline (mg/dl) | −16±57 | −7±60 | ||
Data given as mean±SD or †geometric mean±SD. *P<0.05 compared with baseline. #P<0.05 compared with the control group. Comparisons between groups are made by using logarithmic transformed values. HDL-C, high-density lipoprotein cholesterol; HOMA, homeostasis model of assessment; hsCRP, high-sensitive C-reactive protein; IL-6, interleukin-6; LDL-C, low-density lipoprotein cholesterol; RBP4, retinol-binding protein-4; sICAM-1, soluble intercellular adhesion molecule-1; sVCAM-1, soluble vascular cell adhesion molecule-1; TC, total cholesterol; TNF-α, tumor necrosis factor-α.
Total cholesterol and LDL-C decreased significantly in both groups at 9-month follow-up, but the changes from baseline were similar between the 2 groups (Table 4).
The purposes of this prospective, randomized, single-blinded, investigator-initiated, 9-month follow-up study were to compare the effects of pioglitazone on neointimal volume and circulating microRNA-17,-24,-92a,-126,-145 during 9-month follow-up in type 2 diabetic patients. To the best of our knowledge, this is the first study to demonstrate an association between the use of pioglitazone and increases in microRNA-24, and that the increase in microRNA-24 >0.1715 could be used as a potential predictor of neointimal volume <25 mm3 during 9-month follow-up. Moreover, this study has demonstrated that pioglitazone improved brachial artery endothelial dysfunction in type 2 diabetic patients with severe coronary artery stenosis requiring stenting.
MicroRNAs are a class of endogenous, small, non-coding RNAs that regulate expression of the protein-coding genes via degradation or translational inhibition of their target messenger RNAs.11,21,22 After screening endothelial function-related microRNAs such as microRNA-17,-21,-92a,-126,-145,-24,-26,-143,-155,-423,-1,-10a,-100,-204, and -208a in peripheral blood of patients with (n=5) or without (n=5) in-stent restenosis in our pilot study, high expression of microRNA-17,-24,-92a,-126, and -145 was noted. MicroRNA-17,-92a are expressed in endothelial cells and have been known to control endothelial cell function and atherosclerotic lesion formation.13,15,16 An atheroprotective role of mircoRNA-126 has been shown in apolipoprotein E-deficient mice,17 and microRNA-145 is a novel vascular smooth muscle cell phenotypic marker and modulator that is capable of controlling vascular neointimal lesion formation.12 MicroRNA-24 is known for regulating target proteins involved in cellular survival, apoptosis and cell invasiveness.18,19,23 MicroRNA-24 is known to play critical roles in the pathogenesis of cardiac cell growth and death, smooth muscle cell proliferation and apoptosis, and cardiac fibroblast activation.18,19,23 Expression of microRNA-24 is downregulated in cardiomyocytes after myocardial infarction, and upregulated microRNA-24 could reduce myocardial fibrosis in the infarct border zone.18 Increases in neointimal hyperplasia, which is an indication of smooth muscle cell proliferation, could be identified by decreases in circulating microRNA-24 in type 2 diabetic patients in this study. Although other microRNAs such as microRNA-17,-92a,-126, and -145 were found to be potential indicators of neointimal hyperplasia in our pilot study, the same trends were not seen in the present study, confined to type 2 diabetic patients during 9-month follow-up.
The thiazolidinediones such as pioglitazone interact with the PPAR-γ ligand-binding domain and are known for their hypoglycemic effect, controlling dyslipidemia, and reducing inflammation.9,24,25 The nuclear hormone receptor PPAR-γ is activated by its ligand, and PPAR-γ functions as a transcriptional regulator of multiple genes regulating glucose and lipid metabolism.9,24 In addition to their hypoglycemic effect and correction of dyslipidemia, pioglitazone has anti-atherogenic effects in vascular cells in vitro.26 Treatment with PPAR-γ activators such as pioglitazone and telmisartan reduced inflammatory cells and systemic inflammation after coronary stenting, which was associated with reduction in neointimal volume during the follow-up.9,20,27,28 Early PPAR-γ activation in inflammatory cells with subsequent reduction in systemic inflammation after coronary stenting is critical in the reduction of neointimal volume and atherosclerosis progression, especially in diabetic patients. Pioglitazone is known to interact with the PPAR-γ ligand-binding domain, thereby reducing inflammation, LDL-C and insulin resistance in type 2 diabetic patients.29,30 In this study, total cholesterol and LDL-C decreased in both groups, with no significant differences in changes vs. baseline between groups because atorvastatin 20 mg or rosuvastatin 10 mg was used in 97.2% of study patients (97.2%, n=35 in the pioglitazone group and 97.2%, n=36 in the control group, P=1.000). The beneficial vascular effects of pioglitazone were independent of its improvement of dyslipidemia and insulin resistance in the present study.
There are several possible mechanisms for the significant reduction in neointimal hyperplasia in the pioglitazone group when compared with the control group during follow-up, even though the extent of HOMA index and LDL-C reduction was similar for either treatment during follow-up. The inflammatory markers such as IL-6, TNF-α, and sVCAM-1 decreased significantly in pioglitazone compared with the control, thereby reducing inflammatory condition during follow-up. In this study, adiponectin, which is involved in various metabolic processes and which influences neointimal hyperplasia after coronary stenting, increased significantly only in the pioglitazone group.20 Adiponectin stimulates nitric oxide production in endothelial cells and blunts the secretion of TNF-α from macrophages and the ensuing expression of adhesions on the endothelium, thereby inhibiting several essential steps of neointimal proliferation.31 Adiponectin, in turn, has been shown to decrease cytokine production from macrophages and to interfere with TNF-α signaling, thereby reducing inflammatory conditions after stent implantation.32,33 Increased IL-6 and TNF-α increases the synthesis of CRP from coronary artery smooth muscle cells;34 the CRP produced in coronary artery smooth muscle cells could directly participate in excessive neointimal hyperplasia after coronary stenting. hsCRP, however, decreased in both groups, with no significant differences in changes vs. baseline between groups. The present results suggest that the change in adiponectin, IL-6 and TNF-α is more directly influenced by the pioglitazone-induced activation of PPAR-γ than the change in hsCRP. We therefore assume that production of hsCRP, which is a downstream inflammatory cascade marker, was influenced by a variety of risk factors and medications in this study. Pioglitazone, due to its potent anti-inflammatory effects, also significantly reduced neointimal volume and increased baFMD.
RBP4 is an adipokine that is elevated in serum during insulin-resistant conditions, and genetic deletion of RBP4 increases insulin sensitivity.35 RBP4, in addition to the HOMA index and HbA1c, decreased significantly in both groups, with no significant differences in change vs. baseline between groups (Table 4), because oral hypoglycemic medications and insulin were used at physician discretion to control blood glucose level in this study.
There are a few limitations to the present study. Although this study was adequately powered to compare neointimal volume with OCT between the 2 groups, the number of subjects was not sufficient to draw any conclusions on pioglitazone-associated cardiovascular events during long-term follow-up. Moreover, the exact molecular mechanisms underlying the anti-inflammatory and anti-atherogenic activities of pioglitazone and its association with microRNA-24 remain to be verified. Because this study was confined to type 2 diabetic patients, the present findings should not be extrapolated to all patients after coronary stent implantation.
Type 2 diabetic patients treated with pioglitazone not only benefit from its hypoglycemia- and LDL-C-lowering effects but also from its anti-inflammatory and circulating microRNA-24-increasing effects, which decrease neointimal proliferation and improve endothelial dysfunction.
Circulating microRNA-24 was aberrantly decreased in type 2 diabetic patients with excessive neointimal hyperplasia; therefore, modulation of microRNA-24 expression with a pharmacological approach such as pioglitazone has strong downregulating effects on neointimal proliferation in type 2 diabetic patients. Circulating microRNA-24 could be used as a potential novel biomarker for predicting excessive neointimal hyperplasia in type 2 diabetic patients after coronary stent implantation.
This study was supported by the 2011 grant from the Korea University (K1132211) and by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health and Welfare, Republic of Korea (HI14C0209).
Conflict of Interest: No potential conflict of interest relevant to this article was reported.
Supplementary File 1
Figure S1. Significant negative correlations between neointimal volume and (A) microRNA-24; (B) flow-mediated dilation at 9 months; and (C) plasma adiponectin concentration.
Table S1. Adverse clinical events during 9-month follow-up
Table S2. Angiographic parameters during 9-month follow-up
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-14-0964
