The interaction between CD40 and CD40 ligand (CD40L,

Transcription

1 Soluble CD40 Ligand, Soluble P-Selectin, Interleukin-6, and Tissue Factor in Diabetes Mellitus Relationships to Cardiovascular Disease and Risk Factor Intervention Hoong Sern Lim, MRCP; Andrew D. Blann, PhD; Gregory Y.H. Lip, MD Background High levels of the soluble fragment of CD40 ligand (scd40l) have previously been associated with adverse cardiovascular outcomes. CD40L CD40 interaction has been linked to the pathogenesis of atherothrombotic complications in cardiovascular disease (CVD). We sought to determine whether a package of care of intensified multifactorial cardiovascular risk intervention could reduce indices of platelet activation, inflammation, and coagulation in diabetes and whether patients with overt CVD would derive similar benefit compared with those without. Methods and Results We measured plasma scd40l, soluble P-selectin (sp-sel, an index of platelet activation), interleukin-6 (IL-6, a proinflammatory cytokine), and tissue factor (TF, an initiator of coagulation) in 97 patients with diabetes mellitus (41 with and 56 without overt CVD) and 39 comparable healthy control subjects. Thirty-six patients with and 32 without overt CVD then participated in a package of care of cardiovascular risk intervention over a period of 1 year. Plasma levels of scd40l (P 0.001), sp-sel (P 0.001), IL-6 (P 0.001), and TF (P 0.001) were higher in patients with diabetes than in control subjects, with TF levels highest in patients with overt CVD. Multifactorial intervention was associated with significant reductions in scd40l in both patient groups (both P 0.001), but reductions in sp-sel and TF were seen only in patients without overt CVD. There was no significant change in IL-6 levels in both patient groups. Conclusions Intensive multifactorial risk management can reduce high levels of scd40l but can only partially correct abnormal platelet activation, inflammation, and coagulation in diabetes, particularly in patients with overt CVD. (Circulation. 2004;109: ) Key Words: diabetes mellitus CD40 ligand tissue factor interleukin-6 P-selectin The interaction between CD40 and CD40 ligand (CD40L, or officially, CD154) was initially described in antigen presentation and B and T lymphocyte biology. 1 CD40L has now been described in nonleukocytic cells, including endothelial cells and platelets, suggesting a wider role beyond cellular immunity. Indeed, more recent data implicate this CD40 CD40L system in the pathogenesis of atherothrombotic complications (and prognosis) in cardiovascular disease (CVD) as well as in inflammation and thrombosis. 2 Normally absent from the surface of unstimulated platelets, CD40L is rapidly presented to the platelet surface after platelet activation. 3 Platelet-associated CD40L on exposure to CD40-expressing vascular cells (including endothelial cells) induces the expression of adhesion molecules, the release of inflammatory cytokines (eg, interleukin-6, IL-6), 4 and the procoagulant tissue factor (TF). 5,6 Surface-expressed CD40L is subsequently cleaved to generate a soluble fragment (ie, soluble CD40 ligand [scd40l]). This soluble fragment retains much of its biological activities by virtue of binding to glycoprotein IIb/IIIa and the induction of signaling reactions when bound to receptors. 7 Like soluble P-selectin (sp-sel), circulating scd40l is believed to derive predominantly from activated platelets and hence may reflect platelet activation. 8,9 Therefore, current data in vitro support a link between platelet activation, thrombogenesis, and the inflammatory state. High levels of scd40l, 10,11 IL-6, 12 TF, 13 and sp-sel 14 have previously been shown in patients with diabetes compared with healthy control subjects and may be associated with adverse cardiovascular outcome However, the in vivo relationship(s) between these indices in patients with diabetes, with and without overt CVD, are not clear. Treatment with HMG-CoA reductase inhibitors (statins) and peroxisome proliferator activated receptor- agonist (glitazones) has been reported to reduce plasma scd40l levels, 10,11,19 but the effects on plasma IL-6 have been inconsistent. 20 Indeed, the contemporary approach to the management of patients with diabetes is a package of care that targets elevated blood glucose (typically with sulfonylurea- or insulin-based ther- Received November 17, 2003; revision received February 13, 2004; accepted February 17, From the Haemostasis Thrombosis and Vascular Biology Unit, University Department of Medicine, City Hospital, Dudley Road, Birmingham, England. Correspondence to Professor G.Y.H. Lip, Haemostasis Thrombosis and Vascular Biology Unit, University Department of Medicine, City Hospital, Dudley Road, Birmingham, B18 7QH, England, UK. g.y.h.lip@bham.ac.uk 2004 American Heart Association, Inc. Circulation is available at DOI: /01.CIR

2 Lim et al scd40l and Cardiovascular Risk Intervention 2525 apy), arterial blood pressure, dyslipidemia, and abnormal platelet activation. This strategy reduces microvascular and cardiovascular complications in diabetes, 21 but the corresponding effects on platelet activation, scd40l levels, and the inflammatory and hypercoagulable state in diabetes remain unclear. Also unclear is the extent to which overt CVD itself influences the effect of intervention on these indices in these already high-risk patients. First, we hypothesized a relationship between scd40l and sp-sel, TF, and IL-6 (as indices of platelet activation, thrombogenesis, and the inflammatory state, respectively) in patients with type 2 diabetes mellitus, with the highest levels in those with cardiovascular disease. Second, we hypothesized that a package of care of intensified multifactorial cardiovascular risk intervention would reduce these indices, and third, that the reduction would be more marked in patients without CVD than in patients with overt CVD. We tested our hypotheses in a cross-sectional study of plasma scd40l, sp-sel, IL-6, and TF in patients with type 2 diabetes mellitus compared with healthy controls. Furthermore, we determined the effects of an intervention study (in which patients participated in a package of care of cardiovascular risk intervention over a period of 1 year) on these indices. Methods Ninety-seven patients with type 2 diabetes (diagnosed according to the WHO criteria) were recruited. Of these, 41 had previous ( 3 months) stroke, myocardial infarction, unstable angina, and coronary and peripheral revascularization. All the patients were on antihypertensive and oral hypoglycemic therapy. Sixty-eight patients (of whom 32 did not have clinically overt CVD) consented to participate in a package of care of intensified diabetes and cardiovascular risk management. These patients were offered 3 monthly consultations, providing lifestyle advice, recommending at least three 30-minute sessions of light-to-moderate exercise per week. In addition, their oral hypoglycemic therapy was adjusted to target glycosylated hemoglobin (HbA 1c )of 6.5%. Metformin was started in overweight patients (defined as body mass index [BMI] of 25 kg/m 2 ) and added as second-line agent in lean patients, who received gliclazide modified release as first-line treatment (maximum daily dose of 120 mg). Gliclazide MR was added in overweight patients if hyperglycemia remained uncontrolled on metformin. Insulin therapy was recommended for patients whose HbA 1c remained 7.0% on maximal doses of oral hypoglycemic agents. Blood pressure management followed a similar stepwise approach, to a target of 130/80 mm Hg. All patients were initially commenced on an ACE inhibitor. A diuretic, calcium channel blocker, or -blocker could be added as additional agents. The use of lipid-lowering therapy (HMG-CoA reductase inhibitor or statins) and low-dose aspirin was encouraged in all patients. HbA 1c, blood pressure, lipid profile, and research indices were measured at baseline and 1 year. Waist measurements were taken at the level of the umbilicus. Venous blood was obtained by atraumatic venepuncture into sodium citrate and was immediately centrifuged at 1000g and 4 C for 20 minutes. Plasma was divided into aliquots and stored at 70 C for batch analysis. Data from patients with diabetes were compared with healthy controls recruited from relatives and friends of patients. These subjects were normotensive and without clinical evidence of vascular, neoplastic, metabolic, or inflammatory disease by careful clinical history, examination, and routine laboratory tests. This study was approved by the West Birmingham Local Research Ethics Committee, and informed consent was obtained from all participants. Laboratory We analyzed plasma IL-6, scd40l, sp-sel (all from R&D Systems), and TF (Axis-Shield) by ELISA using commercial kits and reagents. All assays have intra-assay and interassay coefficients of variation of 5% and 10%, respectively. HbA 1c was measured by liquid chromatography (BioRad Variant 2, BioRad). The urinary albumin:creatinine ratio was measured by immunoturbidimetry (I- Laboratory 600 Clinical Chemistry System, Instrumentation Laboratory) from single-void urine samples taken on 2 separate occasions at least 2 weeks apart. The intra-assay and interassay coefficients of variation were also 5% and 10%, respectively. Power Calculations At the time of the study, there were no data comparing differences in scd40l in patients with diabetes with and without overt CVD. We based our power calculations on previous studies 10,11 and hypothesized that differences in levels of scd40l will be of similar magnitude. Therefore, seeking a difference of 0.8 SD, we needed data from a minimum of 35 patients in each of the 3 groups (ie, healthy controls, diabetes free of overt CVD, and diabetes with existing CVD) for P 0.05 and However, we recruited in excess of this figure for additional confidence. Serial pairwise measurements in 30 subjects provided the power to detect a difference of 0.5 SD at P 0.05 and Statistical Analysis Continuous data were subjected to the Anderson-Darling test to determine their distribution. All 4 major research indices (Table 2) were nonnormally distributed and were therefore logarithmically transformed for analysis. The difference between the groups in terms of means and SDs of these log-transformed data indicated that we had sufficient power to test our hypothesis. Nonnormal data, presented as median and interquartile range, were analyzed by the Mann-Whitney U test (2 groups) or the Kruskall-Wallis test (3 groups). Normally distributed data, presented as mean and SD were analyzed by Student s unpaired t test (2 groups) or ANOVA (3 groups). Categorical data were analyzed by the 2 test. Correlations within each group were sought by use of Spearman s method. The significance of any changes in the clinical and biochemical indices with therapy was evaluated with the paired t test or Wilcoxon s signed-rank test for normal and nonnormal data, respectively. Correlation coefficients were computed to assess the association between changes in sp-sel, IL-6, TF, and scd40l with changes in clinical and metabolic parameters. All analyses and power calculations were performed using Minitab 13 (Minitab Inc). Results Ninety-seven patients (41 patients with and 56 without overt CVD) and 39 healthy controls were recruited (Table 1). There were no significant differences in age and the male-to-female ratio. Waist circumference, BMI, systolic blood pressure, and HbA 1c were significantly higher, whereas total and HDL cholesterol were significantly lower in patients compared with controls. BMI was highest in patients without overt CVD. There was no significant difference in the use of statins, aspirin, or other clinical or metabolic parameters between the 2 groups of patients. Plasma scd40l, sp-sel, IL-6, and TF Levels in Patients and Controls As expected from existing data, plasma levels of scd40l, sp-sel, IL-6, and TF were significantly higher in patients with diabetes than controls, with TF levels being highest in patients with overt CVD (Table 2, Figure).

3 2526 Circulation June 1, 2004 TABLE 1. Characteristics of Patients and Controls Controls (n 39) Diabetes Without Overt CVD (n 56) Diabetes With Overt CVD (n 41) P Age, y Male, n Waist, cm BMI, kg/m * SBP, mm Hg DBP, mm Hg HbA 1c,% Cholesterol, mmol/l Triglycerides, mmol/l 1.6 ( ) 1.6 ( ) 1.8 ( ) HDL, mmol/l Urine albumin:creatinine ratio, mg/mmol 1.4 ( ) 1.4 ( ) 0.62 Statin use, n BMI indicataes body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; sp-sel, soluble P-selectin; IL-6, interleukin-6; scd40l, soluble CD40 ligand. *Significantly higher in patients without vs patients with overt CVD and controls but no significant difference between patients with overt CVD and controls on ANOVA and Tukey s post hoc analysis. Significantly higher in patients vs controls but no significant difference between patient groups on ANOVA and Tukey s post hoc analysis. Higher in controls vs patients but no significant difference between patient groups on ANOVA and Tukey s post hoc analysis. Correlations In the cohort of patients with diabetes, with or without overt CVD, scd40l, sp-sel, IL-6, and TF failed to correlate significantly with clinical parameters (age, blood pressure, waist measurement, or BMI), indices of metabolic control (lipid profile and glycohemoglobin, HbA 1c ), or the degree of microalbuminuria (diabetic microvascular complications) (some data not shown). However, scd40l correlated significantly with IL-6 and TF in patients with and without overt CVD (Table 3). Effects of Intervention In the 32 patients with diabetes but free of overt CVD, statin use increased and glycemic control improved significantly at the end of the year, with corresponding reductions in total cholesterol and triglycerides. There were no significant differences in mean body weight, blood pressure, and HDL cholesterol. These changes were associated with significant reductions in plasma scd40l to a median of 60% of preintervention levels, sp-sel to 92%, and TF to 55% but not plasma IL-6 levels (Table 4). Statin and aspirin use increased significantly in the 36 patients with diabetes and overt CVD at 1 year compared with baseline. Total cholesterol and glycemia improved significantly, but there was no significant difference in body weight, blood pressure, HDL cholesterol, and triglycerides. Plasma scd40l level fell significantly (to 47% of preintervention levels), but the changes in sp-sel, TF, and IL-6 were not significant (Table 4). Influence of Changes in Metabolic Parameters We performed additional analyses to assess whether the changes in scd40l, sp-sel, and TF were associated with changes in the metabolic parameters (total cholesterol, triglycerides, and HbA 1c ). In patients without overt CVD, the changes in scd40l, sp-sel, and TF did not correlate with the TABLE 2. Plasma scd40l, sp-sel, IL-6, and TF in Patients With Diabetes (With and Without Overt CVD) and Controls Controls (n 39) Diabetes Without Overt CVD (n 56) Diabetes With Overt CVD (n 41) P scd40l, pg/ml 30 (10 175) 520 ( ) 700 ( ) 0.001* sp-sel, ng/ml 65 (50 85) 95 (80 115) 100 (85 115) 0.001* IL-6, pg/ml 1.0 ( ) 1.9 ( ) 2.6 ( ) 0.001* TF, pg/ml 300 (11 925) 510 ( ) 1300 ( ) TF indicates tissue factor; IL-6, interleukin-6; sp-sel, soluble P-selectin; scd40l, soluble CD40 ligand. *Significantly higher in patients vs controls but no significant difference between patient groups on ANOVA and Tukey s post hoc analysis (log-transformed data). Higher in patients vs controls and higher in patients with vs patients without overt CVD on ANOVA and Tukey s post hoc analysis (log-transformed data).

4 Lim et al scd40l and Cardiovascular Risk Intervention 2527 TABLE 3. Correlations in Patients Without and With Overt CVD Without overt CVD UACr HbA 1c sp-sel IL-6 TF Plasma soluble CD40 ligand levels in patients and controls. reductions in total cholesterol, triglycerides, or HbA 1c. Changes in sp-sel and TF did not correlate with the decrease in scd40l in patients without overt CVD. Similarly, in patients with overt CVD, the decrease in plasma scd40l did not correlate with changes in total cholesterol and HbA 1c (all P NS, data not shown). Discussion The interaction between CD40 and CD40L, by inducing proinflammatory cytokines, metalloproteinases, and the procoagulant TF, may accelerate atherogenesis, weaken atheromatous plaques, and promote thrombotic occlusion of the artery. 2 Circulating scd40l, believed to derive predominantly from activated platelets, has been shown to possess biological activity. 3 Accordingly, high levels of scd40l have been shown to predict future cardiovascular events in clinically healthy women, 18 as well as recurrent events and death in patients presenting with acute coronary syndromes. 22 Consistent with previous reports, we have confirmed higher levels of scd40l in patients with diabetes compared with controls. 10,11 In addition, we have demonstrated strong correlations between scd40l and both IL-6 and TF, hence extending previous in vitro findings 4,5 and supporting a link between the CD40 CD40L system and the underlying inflammatory and hypercoagulable state in these patients in vivo. However, we did not find any significant correlations between scd40l and total cholesterol or sp-sel (a recognized marker of platelet activation 9 ). This is in contrast to the study by Cipollone et al, 19 who demonstrated significant correlations with total and LDL cholesterol and sp-sel. Nevertheless, the latter study consisted of hypercholesterolemic, nondiabetic patients who were not taking antiplatelet agents, a study population quite distinct from that in our study, which may account for the observed differences. The use of the peroxisome proliferator activated receptor agonists troglitazone 11 and rosiglitazone, 10 as well as statin 19 therapy, is associated with reductions in scd40l levels in patients with diabetes and hypercholesterolemia, respectively. In the present study, we have shown significant reductions in circulating scd40l in patients with diabetes, with and without clinically overt CVD, following a package HbA 1c r P sp-sel r P IL-6 r P TF r P scd40l r P With overt CVD HbA 1c r P sp-sel r P IL-6 r P TF r P scd40l r P TF indicates tissue factor; IL-6, interleukin-6; sp-sel, soluble P-selectin; scd40l, soluble CD40 ligand; UACr, urine albumin:creatinine ratio. of care consisting of ACE inhibitor based blood pressure control, statin-based lipid lowering, sulfonylurea-based glycemic control, and antiplatelet therapy. Gaede et al 21 have recently described significant reductions in microvascular and cardiovascular complications with similar target-driven multifactorial intervention, but we accept that any suggestion that these findings may be related to reductions in scd40l levels remains speculative at present. Of note, the lack of significant correlations between scd40l (and sp-sel) and glycemia, blood pressure, and cholesterol, or indeed the changes in scd40l (and sp-sel) with changes in glycemic and lipid parameters, suggests alternative mechanism(s) for the elevated scd40l and abnormal platelet activation in patients with diabetes. The presence of clinically overt CVD, although not associated with significant difference in plasma scd40l, appears to be associated with higher levels of TF and IL-6. Further-

5 2528 Circulation June 1, 2004 TABLE 4. Changes in Clinical, Metabolic, and Biochemical Parameters With a Package of Care of Intensified Diabetes and Cardiovascular Risk Management in Patients With Diabetes Without and With Overt CVD Baseline 1 Year P Without overt CVD (n 32) SBP, mm Hg DBP, mm Hg Weight, kg HbA 1c,% Total cholesterol, mmol/l Triglycerides, mmol/l 1.8 ( ) 1.5 ( ) HDL, mmol/l Statin use, n Aspirin use, n sp-sel, ng/ml IL-6, pg/ml 1.0 ( ) 1.0 ( ) scd40l, pg/ml 480 ( ) 285 ( ) TF, pg/ml 400 ( ) 220 ( ) With overt CVD (n 36) SBP, mm Hg DBP, mm Hg Weight, kg HbA 1c,% Total cholesterol, mmol/l Triglycerides, mmol/l 1.6 ( ) 1.6 ( ) HDL, mmol/l Statin use, n Aspirin use, n sp-sel, ng/ml IL-6, pg/ml 1.0 ( ) 1.0 ( ) scd40l, pg/ml 530 ( ) 250 (21 788) TF, pg/ml 950 ( ) 900 ( ) SBP indicates systolic blood pressure; DBP, diastolic blood pressure; sp-sel, soluble P-selectin; IL-6, interleukin-6; scd40l, soluble CD40 ligand; TF, tissue factor. more, patients with overt CVD did not appear to derive similar reductions in plasma sp-sel and TF compared with patients without overt CVD. High levels of sp-sel, TF, and IL-6 have been associated with adverse cardiovascular outcomes and the development of coronary events in subjects without clinically established CVD. 15,16,17 This may explain the risk continuum observed in healthy subjects to patients with diabetes free of overt CVD to patients with established CVD and supports early intensive multifactorial intervention. In conclusion, we have shown that the elevated plasma scd40l in patients with diabetes mellitus, with and without overt CVD, can be reduced by intensive application of a package of care of a contemporary target-driven, multifactorial cardiovascular risk intervention strategy. However, the lack of impact of this intervention strategy on the underlying inflammatory and hypercoagulable state, particularly in patients with clinically overt CVD, suggests the need for earlier and more aggressive cardiovascular risk interventions in diabetes. Acknowledgment We acknowledge the support of the Sandwell and West Birmingham Hospitals NHS Trust Research and Development Programme for support of the Hemostasis, Thrombosis, and Vascular Biology Unit. References 1. Grewal IS, Flavell RA. CD40 and CD154 in cell-mediated immunity. Annu Rev Immunol. 1998;16: Schonbeck U, Libby P. CD40 signaling and plaque instability. Circ Res. 2001;89: Henn V, Slupsky JR, Grafe M, et al. CD40 ligand on activated platelets triggers an inflammatory reaction of endothelial cells. Nature. 1998;391: Deschanet J, Grosset C, Taupin JL, et al. CD40 ligand stimulates proinflammatory cytokine production by human endothelial cells. J Immunol. 1997;159: Slupsky JR, Kalbas M, Willuweit A, et al. Activated platelets induce tissue factor expression on human umbilical vein endothelial cells by ligation of CD40. Thromb Haemost. 1998;80: Zhou L, Stordeur P, de Lavareille A, et al. CD40 engagement on endothelial cells promotes tissue factor-dependent procoagulant activity. Thromb Haemost. 1998;79: Andre P, Prasad KS, Denis CV, et al. CD40L stabilises arterial thrombi by a beta 3 integrin-dependent mechanism. Nat Med. 2002;8: Henn V, Steinbach S, Buchner K, et al. The inflammatory action of CD40 ligand (CD154) expressed on activated human platelets is temporarily limited by coexpressed CD40. Blood. 2001;98: Blann AD, Lip GY. Hypothesis: is soluble P-selectin a new marker of platelet activation? Atherosclerosis. 1997;128: Marx N, Imhof A, Froehlich J, et al. Effect of rosiglitazone treatment on soluble CD40L in patients with type-2 diabetes and coronary artery disease. Circulation. 2003;107: Varo N, Vicent D, Libby P, et al. Elevated plasma levels of the atherogenic mediator soluble CD40 ligand in diabetic patients: a novel target of thiazolidinediones. Circulation. 2003;107: Bermudez EA, Rifai N, Buring J, et al. Interrelationships among circulating interleukin-6, C-reactive protein, and traditional cardiovascular risk factors in women. Arterioscler Thromb Vasc Biol. 2002;22: Zumbach M, Hofmann M, Borcea V, et al. Tissue factor antigen is elevated in patients with microvascular complications of diabetes mellitus. Exp Clin Endocrinol Diabetes. 1997;105: Steiner M, Reinhardt KM, Krammer B, et al. Increased levels of soluble adhesion molecules in type 2 (non-insulin dependent) diabetes mellitus are independent of glycaemic control. Thromb Haemost. 1994;72: Soejima H, Ogawa H, Yasue H, et al. Heightened tissue factor associated with tissue factor pathway inhibitor and prognosis in patients with unstable angina. Circulation. 1999;99: Ridker PM, Rifai N, Stampfer MJ, et al. Plasma concentration of interleukin-6 and the risk of future myocardial infarction among apparently healthy men. Circulation. 2000;101: Ridker PM, Buring JE, Rifai N. Soluble P-selectin and the risk of future cardiovascular events. Circulation. 2001;103: Schonbeck U, Varo N, Libby P. Soluble CD40L and cardiovascular risk in women. Circulation. 2001;104: Cipollone F, Mezzetti A, Porreca E, et al. Association between enhanced soluble CD40L and prothrombotic state in hypercholesterolemia: effects of statin therapy. Circulation. 2002;106: Kinlay S, Schwartz GG, Olsson AG, et al. High-dose atorvastatin enhances the decline in inflammatory markers in patients with acute coronary syndromes in the MIRACL study. Circulation. 2003;108: Gaede P, Vedel P, Larsen, et al. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med 2003; 348: Heeschen C, Dimmeler S, Hamm CW, et al. Soluble CD40 ligand in acute coronary syndromes. N Engl J Med. 2003;348: