Antidiabetic PPARγ-Activator Rosiglitazone Reduces MMP-9 Serum Levels in Type 2 Diabetic Patients With Coronary Artery Disease
Background— Matrix metalloproteinases (MMPs) are critically involved in the development of unstable plaques. Although arteriosclerotic lesions in patients with diabetes mellitus are more unstable than those of nondiabetic subjects, nothing is known about serum levels of MMPs in these patients or about mechanisms to modulate MMP levels. We investigated MMP levels in diabetic and nondiabetic coronary artery disease (CAD) patients and performed a clinical trial to assess the effect of the PPARγ-activating, antidiabetic thiazolidinedione rosiglitazone on MMP levels in diabetic CAD patients.
Methods and Results— In CAD patients, MMP-2, -8, and -9 serum levels were significantly higher in type 2 diabetic subjects compared with age-, sex-, and body mass index–matched nondiabetics. Thirty-nine diabetic patients with CAD were randomized to receive rosiglitazone 4 mg (twice daily) or placebo for 12 weeks. Rosiglitazone treatment, but not placebo, significantly reduced MMP-9 levels already after 2 weeks by −19.6% (−38.3% to 8.6%, P<0.05), and levels remained suppressed until the end of the study. In addition, rosiglitazone significantly decreased serum amyloid A (SAA) and tumor necrosis factor-α levels.
Conclusion— MMP-9 levels are increased in type 2 diabetic patients with CAD, and treatment of these patients with the antidiabetic PPARγ-activator rosiglitazone significantly reduces MMP-9, tumor necrosis factor-α, and SAA serum levels. These data support anti-inflammatory and potential antiatherogenic effects of thiazolidinediones.
Patients with diabetes mellitus have an increased risk of developing extensive arteriosclerosis with its sequelae, unstable angina pectoris, and acute myocardial infarction.1 In the last decade, experimental data have illuminated the role of inflammation in atherogenesis while clinical studies have shown that this emerging concept of inflammation in arteriosclerosis applies directly to human patients.2 During plaque development, vascular cells, such as monocytes/macrophages, T-cells, endothelial cells, and vascular smooth muscle cells, release inflammatory mediators like cytokines or chemokines, thus orchestrating an ongoing inflammatory response in the vessel wall.3 Prospective epidemiological studies have found increased cardiovascular risk with elevated basal levels of some of these same mediators. In particular, cytokines such as interleukin (IL)-6 or tumor necrosis factor-α (TNF-α), soluble adhesion molecules, and downstream acute phase reactants such as C-reactive protein (CRP), fibrinogen, and serum amyloid A (SAA) have been shown to predict the risk of cardiovascular events.2 In addition to these markers, matrix-degrading matrix metalloproteinases (MMPs), such as MMP-2, -8, and -9, have been implicated in plaque rupture through their capacity to thin the protecting fibrous cap of the plaque, thus rendering it more vulnerable.4,5⇓ In fact, MMP-9 levels are elevated in patients with unstable plaques,6 and MMP-2 as well as MMP-9 levels increase in acute myocardial infarction, suggesting that these MMPs contribute to acute coronary syndromes.7 Arteriosclerotic lesions in patients with diabetes mellitus are usually more vulnerable than lesions from nondiabetic subjects, thus potentially explaining their increased cardiovascular risk,8 but hitherto, nothing is known about the role of MMP serum levels in diabetic patients with coronary artery disease (CAD). Given the increased propensity of diabetic patients to develop macrovascular events, therapeutical strategies that limit inflammation in the vessel wall and reduce serum levels of inflammatory surrogate parameters as well as MMPs might be a promising tool to influence vascular disease in this high-risk population.
Recent experimental data suggest that a novel group of antidiabetic agents, thiazolidinediones (TZDs, Glitazones), like rosiglitazone or pioglitazone, might–in addition to their metabolic effects–exhibit anti-inflammatory properties in the vessel wall. These agents lower blood glucose levels by increasing insulin sensitivity in peripheral organs and influence lipid metabolism in treated patients. On a molecular level, TZDs activate the nuclear transcription factor peroxisome proliferator-activated receptor gamma (PPARγ), thus inducing or inhibiting the expression of various target genes.9 In vascular cells, PPARγ-activating TZDs inhibit monocyte cytokine and MMP-9 secretion,10–12⇓⇓ reduce endothelial chemokine and adhesion molecule expression, and limit smooth muscle cell and T-cell activation (reviewed by Marx13 and Marx et al14). In addition, TZDs have been shown to decrease lesion size in various animal models of arteriosclerosis.15,16⇓ Recent work by Haffner et al17 demonstrated reduced serum levels of CRP, MMP-9, and white blood cell count (WBC) after 26 weeks of treatment with rosiglitazone compared with placebo. However, nothing is known about the effect of TZD treatment on other MMPs, considered to be of importance in unstable plaques, nor on the time dependency of these pleiotropic effects.
Therefore, we compared MMP levels in CAD patients with type 2 diabetes with those of nondiabetic subjects and performed a randomized, placebo-controlled trial to examine the effect of rosiglitazone treatment on serum levels of MMPs and other inflammatory markers of arteriosclerosis.
Patients and Study Design
Patients were recruited from the Department of Internal Medicine II at the University of Ulm, Germany. Patients, aged 30 to 75 years, with established type 2 diabetes for at least six month (ADA criteria 1997), who were on oral antidiabetic treatment only (metformin, sulfonylurea) and had CAD (>50% stenosis by angiography), were included. Exclusion criteria were as follows: cardiogenic shock, unstable angina or myocardial infarction during the preceding 4 weeks, impaired liver function, insulin treatment, renal failure requiring hemodialysis, cancer, systemic inflammatory disease, previous TZD treatment, and recent major surgery or illness.
The control group consisted of 21 nondiabetic CAD patients with identical inclusion and exclusion criteria besides the absence of diabetes mellitus. The study protocol was approved by the local Ethics Committee, and all patients gave written, informed consent.
For the clinical trial, the diabetic patients were randomized to receive either rosiglitazone 4 mg twice daily or placebo on top of individual, conventional treatment for 12 weeks. One patient in the placebo group stopped treatment after 10 days because of nausea. Blood samples were taken before cardiac catheterization as well as after 2, 6, and 12 weeks for measurements of metabolic parameters and inflammatory biomarkers. In the control group, blood samples were also obtained before angiography.
MMP-2, MMP-3, MMP-8, MMP-9, TNF-α, IL-6, soluble-vascular adhesion molecule-1 (sVCAM), soluble-intercellular adhesion molecule-1 (sICAM-1), and soluble-E-Selectin were determined by ELISA (R&D Systems) according to the manufacturers’ protocol. Fibrinogen, SAA, and CRP were measured as previously described.18
Differences between groups were analyzed by using the Mann-Whitney U test. Differences between treatment time-points were calculated by using Friedman RM ANOVA or one-way repeated measurement ANOVA followed by the appropriate post hoc test. Skewed data were reported as median (interquartile range), all other data as mean±SEM. Spearman rank correlation was used to analyze correlation between parameters. A P value <0.05 was regarded as significant.
CAD Patients With Type 2 Diabetes Mellitus Exhibit Higher MMP-2, -8, and -9 Serum Levels Than Nondiabetic CAD Patients
Diabetic and nondiabetic CAD patients did not significantly differ in sex distribution, age, or body mass index (BMI) (Table 1). MMP serum levels in patients were in the range of previously reported values for the assays used. In nondiabetic patients, MMP-2, -8, and -9 serum levels were 1052 ng/mL (879.6 to 1189), 2.3 ng/mL (0–5.4), and 297.8 ng/mL (258.9 to 401.5), respectively, while levels in type 2 diabetic patients were significantly higher at 1291 ng/mL (1041.3 to 1590.5, P<0.01), 4.7 ng/mL (3.2 to 6.6, P<0.05), and 570.5 ng/mL (427.2 to 747.3, P<0.01) for each of the MMPs (Figure 1). MMP serum levels in diabetic patients did not correlate with the duration of diabetes (MMP-2, r=-0.030; MMP-8, r=0.139; MMP-9, 0.038; P=n.s.). Consistent with previous studies,19,20⇓ we found increased levels of sVCAM in patients with diabetes (1071 ng/mL [862.5 to 1414] vs 735 ng/mL [583.7 to 951.3] in nondiabetic patients; P<0.001). No difference was seen for serum levels of MMP-3, CRP, SAA, TNF-α, IL-6, sICAM, and sE-selectin (Table 1).
Rosiglitazone Treatment Reduces MMP-9 Serum Levels in Type 2 Diabetic Patients With CAD
To examine the effect of treatment with the antidiabetic PPARγ-activator rosiglitazone on MMP-2, -8, and -9 levels, a randomized, placebo-controlled, single-blinded trial was performed in type 2 diabetic patients with angiographically proven CAD. Patients of both groups did not significantly differ in any of the baseline characteristics, besides use of metformin monotherapy (Table 2), nor in any of the parameters determined (Table 2). Rosiglitazone treatment, but not placebo, decreased HbA1c levels significantly from 7.5±0.3% to 7.0±0.3% (P<0.01), demonstrating that rosiglitazone exhibited its expected metabolic effects (Figure 2, right panel). Fasting blood glucose levels showed a trend to decreased values in the rosiglitazone group after 12 weeks from 132±14 to 126±7 mg/dL, but this difference did not reach statistical significance (Figure 2, left panel). MMP-9 serum levels, although lower in the placebo group, were not significantly different between the two groups at baseline (P=0.12). Rosiglitazone significantly reduced MMP-9 serum levels already within the first 2 weeks by −19.6% (−38.3% to 8.6%, P<0.05), while no such effect was observed in the placebo group. MMP-9 levels remained suppressed until the end of the study (reduction week 6, −29.1% [−47.3% to 2.5%] and week 12, −24.1% [−37.5% to 11.8%]; both P<0.05 compared with baseline). No significant effects were seen on MMP-2 and -8 serum levels in any group (Figure 3).
Rosiglitazone Treatment Reduces Inflammatory Serum Markers of Arteriosclerosis in Type 2 Diabetic Patients With CAD
To examine whether rosiglitazone’s reduction of MMP-9 levels might extend to other biomarkers of arteriosclerosis, we examined the effect of rosiglitazone on established serum levels of inflammatory markers, such as CRP, SAA, fibrinogen, TNF-α, and IL-6, as well as on serum levels of soluble adhesion molecules, such as sVCAM, sICAM, and sE-selectin. Rosiglitazone treatment significantly reduced SAA levels within 2 weeks, and the effect was preserved till the end of the study. In addition, TNF-α serum levels were significantly reduced in the rosiglitazone group after 6 weeks, further decreasing after 12 weeks (Figure 4). Serum levels of MMP-9, SAA, and TNF-α did not significantly decrease below the levels found in nondiabetic patients (data not shown). Changes in SAA serum levels after 12 weeks of rosiglitazone treatment correlated significantly with changes in fasting blood glucose levels, while no significant correlation was seen between changes of TNF-α and MMP-9, SAA, and MMP-9, or SAA and TNF-α serum levels. Moreover, changes in MMP-9 and TNF-α did not significantly correlate with metabolic changes (Table 3). In addition, rosiglitazone but not placebo showed a trend to decreased serum levels of CRP (reduction after 2 weeks, −24.1% [−59.6% to 23.5%]; after 6 weeks, −37.2% [−63.3% to 0.2%]; and after 12 weeks, −22.4% [−72.1% to 19.6%]), but did not reach statistical significance at any time point (Figure 4). No significant effect was seen on fibrinogen, IL-6, sICAM, sVCAM, or sE-selectin in both treatment groups. In the placebo group, there was no significant reduction in any of the parameters measured (data not shown).
The present study demonstrates that type 2 diabetic patients with CAD exhibit higher MMP-2, -8, and -9 serum levels than nondiabetic CAD patient, and that treatment with the antidiabetic PPARγ-activator rosiglitazone significantly reduces levels of MMP-9, TNF-α, and SAA in diabetic CAD patients.
MMP-2, -8, and -9 are matrix-degrading enzymes critically involved in destabilization of arteriosclerotic lesions, with MMP-9 being more abundantly expressed in unstable than in stable plaques.21 Previous work has established that lesions from diabetic patients are more unstable than those from nondiabetic subjects, thus potentially explaining the increased cardiovascular risk of these patients.8 Our finding of enhanced MMP-2, -8, and -9 levels in diabetic patients with CAD compared with matched nondiabetic subjects might thus potentially reflect the presence of more unstable lesions in these patients. However, a recent study by Portik-Dobos et al22 found decreased MMP-9 and MMP-2 expression in vessels from diabetic compared with nondiabetic patients. Still, the authors of this study examined internal mammaria artery specimens, which were considered relatively healthy and did not address whether the vessel sections in each group contained stable and/or unstable plaques. If plaques in the vessels from diabetic patients were more vulnerable, as previously described by Moreno and colleagues,8 one would expect higher MMP expression in these lesions. However, to date there is no evidence that MMP serum levels mirror MMP expression in the vessel wall, especially given that matrix degradation might be a common process in different tissues in diabetic patients. As such, MMP levels might potentially depend on the diabetes stage, although we did not find a significant correlation between diabetes duration and MMP serum levels. Still, further studies are needed to examine the potential prognostic value of MMP-2, -8, and -9 levels in diabetic patients.
The reduction of MMP-9 serum levels on rosiglitazone treatment is in accordance with a recent study by Haffner and colleagues,17 showing significantly reduced MMP-9 levels after 26 weeks of treatment with rosiglitazone in a much larger population of patients with type 2 diabetes mellitus. The present study extends our knowledge on TZDs effects on MMP-9 levels by demonstrating significant reduction of MMP-9 serum levels by rosiglitazone already after 2 weeks of treatment, an effect persisting over the 12-week study period. Interestingly, the extent of MMP-9 reduction seen after the shorter treatment period in our study was similar to the reduction reported after 6 months of rosiglitazone treatment in the previous study at the same dosage. Moreover, former studies have shown that rosiglitazone exhibits its maximal glucose-lowering effects after 8 to 12 weeks.23 This difference strongly suggests that rosiglitazone might directly affect MMP-9 levels independent of its metabolic action. Although TNF-α is known to induce MMP-9 expression, we did not find a significant correlation between changes of these two markers, making reduced TNF-α levels an unlikely explanation for the decrease in MMP-9.
The data obtained here expand our previous in vitro observation that troglitazone, another PPARγ-activating TZD, reduces MMP-9 expression in human macrophages and vascular smooth muscle cells.12,24⇓ In these cells, TZDs had no effect on MMP-2 expression, similar to the lack of an effect of rosiglitazone on MMP-2 serum levels in this study. The in vitro studies did not test MMP-8. Given that MMP-9 levels are elevated in patients with acute coronary syndromes and in those with unstable plaques, reduction of MMP-9 levels by PPARγ-activating TZDs, as seen here, might potentially be interpreted as protective and antiatherogenic properties of these agents, as previously suggested in animal models of arteriosclerosis.15,16⇓ Although rosiglitazone treatment did not significantly reduce serum levels of MMP-9 below the levels of nondiabetic patients, the significant decrease seen in diabetic CAD patients may indicate pleiotropic effects of TZDs on cardiovascular risk markers. However, as mentioned above, further work has to establish the role of MMP serum levels for cardiovascular risk and the importance of MMP lowering for the incidence of macrovascular events.
In addition to the effects on MMP-9, we found a significant reduction of TNF-α and SAA levels by rosiglitazone treatment. These findings may underscore the anti-inflammatory properties of TZDs in high-risk patients, because elevated levels of both markers have been linked to cardiovascular events.2 The effect of rosiglitazone on SAA serum levels is in accordance with a previous study demonstrating a reduction of SAA on 16-week treatment with troglitazone.25 Our study extends the understanding of effects of TZDs on SAA by demonstrating an early reduction of SAA levels after 2 weeks, again suggesting an effect independent of the metabolic action of these agents. SAA is an acute-phase reactant induced by cytokines such as IL-1β and IL-6. Interestingly, we did not find a significant effect of rosiglitazone on IL-6 levels, similar to the study by Haffner and colleagues,17 suggesting that IL-6 levels might not be the best reflection of cytokine levels that bathe the liver. Still, the mechanism through which rosiglitazone reduces SAA levels remains to be elucidated. For CRP, another inflammatory biomarker for cardiovascular events, we found a trend toward reduced levels, but the effects were not significant, most likely due to the small number of patients. However, in a larger population, rosiglitazone has been shown to significantly lower CRP levels.17
In our population, rosiglitazone had no significant effect on levels of the soluble adhesion molecules sICAM, sVCAM, and sE-selectin. Previous in vitro data suggested an effect of TZDs on endothelial adhesion molecule expression,26 and sE-selectin levels have been shown to decrease on troglitazone treatment.27,28⇓ This discrepancy with our data might be due to differences in the molecular structure of these agents, especially because troglitazone, but not rosiglitazone, contains a vitamin E moiety, which has previously been linked to the reduction of endothelial adhesion molecule expression.28 In addition, given the small number of patients, we cannot exclude the possibility that our study might be underpowered to correctly assess all markers measured, and therefore, further studies with TZDs in diabetic patients should include the evaluation of these biomarkers.
Taken together, our study suggests pleiotropic effects of rosiglitazone on inflammatory biomarkers of arteriosclerosis in CAD patients with type 2 diabetes, bolstering the hypothesis that PPARγ-activating TZDs, in addition to their metabolic effects, might be protective in the vessel wall. Still, large trials with clinical endpoints are needed to adequately address this hypothesis.
This work was supported by grants of the Else-Kröner Fresenius Stiftung and the Deutsche Forschungsgemeinschaft to Dr Nikolaus Marx (MA 2047/2-2; SFB 451). The Department of Internal Medicine II, University of Ulm has received an unrestricted grant from SmithKline Beecham, Munich, Germany.
Received October 4, 2002; revision accepted December 4, 2002.
- ↵Libby P, Ridker PM, Maseri A. Inflammation and atherosclerosis. Circulation. 2002; 105: 1135–1143.
- ↵Dollery CM, McEwan JR, Henney AM. Matrix metalloproteinases and cardiovascular disease. Circ Res. 1995; 77: 863–868.
- ↵Moreno PR, Murcia AM, Palacios IF. Coronary composition and macrophage infiltration in atherectomy specimens from patients with diabetes mellitus. Circulation. 2000; 102: 2180–2184.
- ↵Marx N, Sukhova G, Murphy C, Libby P, Plutzky J. Macrophages in human atheroma contain PPARgamma: differentiation-dependent peroxisomal proliferator-activated receptor gamma(PPARgamma) expression and reduction of MMP-9 activity through PPARgamma activation in mononuclear phagocytes in vitro. Am J Pathol. 1998; 153: 17–23.
- ↵Collins AR, Meehan WP, Kintscher U, Jackson S, Wakino S, Noh G, Palinski W, Hsueh WA, Law RE. Troglitazone inhibits formation of early atherosclerotic lesions in diabetic and nondiabetic low density lipoprotein receptor-deficient mice. Arterioscler Thromb Vasc Biol. 2001; 21: 365–371.
- ↵Haffner SM, Greenberg AS, Weston WM, Chen H, Williams K, Freed MI. Effect of rosiglitazone treatment on nontraditional markers of cardiovascular disease in patients with type 2 diabetes mellitus. Circulation. 2002; 106: 679–684.
- ↵Koenig W, Rothenbacher D, Hoffmeister A, Griesshammer M, Brenner H. Plasma fibrin D-dimer levels and risk of stable coronary artery disease: results of a large case-control study. Arterioscler Thromb Vasc Biol. 2001; 21: 1701–1705.
- ↵Albertini JP, Valensi P, Lormeau B, Aurousseau MH, Ferriere F, Attali JR, Gattegno L. Elevated concentrations of soluble E-selectin and vascular cell adhesion molecule-1 in NIDDM: effect of intensive insulin treatment. Diabetes Care. 1998; 21: 1008–1013.
- ↵Otsuki M, Hashimoto K, Morimoto Y, Kishimoto T, Kasayama S. Circulating vascular cell adhesion molecule-1 (VCAM-1) in atherosclerotic NIDDM patients. Diabetes. 1997; 46: 2096–2101.
- ↵Loftus IM, Naylor AR, Goodall S, Crowther M, Jones L, Bell PR, Thompson MM. Increased matrix metalloproteinase-9 activity in unstable carotid plaques: a potential role in acute plaque disruption. Stroke. 2000; 31: 40–47.
- ↵Portik-Dobos V, Anstadt MP, Hutchinson J, Bannan M, Ergul A. Evidence for a matrix metalloproteinase induction/activation system in arterial vasculature and decreased synsthesis and activity in diabetes. Diabetes. 2002; 51: 3063–3068.
- ↵Raskin P, Dole JF, Rappaport EB. Rosiglitazone (RSG) improves glycemic control in porrly controlled insulin-treated type II diabetics (T2D). Diabetes. 1999; 49: 404.
- ↵Marx N, Schönbeck U, Lazar MA, Libby P, Plutzky J. Peroxisome proliferator activated receptor gamma activators inhibit gene expression and migration in human vascular smooth muscle cells. Circ Res. 1998; 83: 1097–1103.
- ↵Pasceri V, Wu HD, Willerson JT, Yeh ET. Modulation of vascular inflammation in vitro and in vivo by peroxisome proliferator-activated receptor-gamma activators. Circulation. 2000; 101: 235–238.
- ↵Cominacini L, Garbin U, Fratta Pasini A, Campagnola M, Davoli A, Foot E, Sighieri G, Sironi AM, Lo Cascio V, Ferrannini E. Troglitazone reduces LDL oxidation and lowers plasma E-selectin concentration in NIDDM patients. Diabetes. 1998; 47: 130–133.