ABSTRACT: Background. The prevalence of diabetes mellitus (DM) and ischemic heart disease is increasing. Moreover, patients with DM experiencing an acute coronary syndrome (ACS) have an increased risk of adverse outcomes after revascularization compared to non-diabetics. Data have suggested that the glycoprotein IIb/IIIa inhibitor abciximab might be more efficient in diabetics than in those without DM. Methods and Results. We evaluated the effect of abciximab in patients with DM and ACS from our percutaneous coronary intervention (PCI) registry. Among 5,003 patients with ACS who underwent PCI, 629 had DM. Patients were followed for up to 3 years with regard to mortality, myocardial infarction (MI) and target vessel revascularization (TVR). Despite a more severe risk profile, adjusted analyses revealed a marked reduction in TVR (hazard ratio [HR], 0.30; confidence interval [CI], 0.14–0.63; p = 0.002), mortality (HR, 0.53; CI, 0.28–0.97; p = 0.04) and the combined endpoint, also including MI (HR, 0.53; CI, 0.35–0.79; p = 0.002) in the DM patients who received abciximab compared to those who did not, resulting in a risk of reaching the endpoints at levels similar to the risk in patients without DM. The reduction in MI was not significant. Conclusion. Our findings suggest that abciximab administered to ACS patients with DM during PCI reduces mortality and the need for TVR to rates similar to those seen in patients without DM and far below the risk in DM patients who do not receive abciximab.
J INVASIVE CARDIOL 2011;23:21–26
Diabetes mellitus (DM), which is becoming more prevalent, is known to be a major risk factor for development of coronary artery disease (CAD). Moreover, diabetic patients with established CAD, e.g., acute coronary syndrome (ACS), have a worse prognosis compared to patients without diabetes. Even though the prognosis with regard to mortality, reinfarction and need for target vessel revascularization (TVR) in general has improved with advances in percutaneous coronary intervention (PCI) devices, new stent types and potent antithrombotic treatment, DM is still associated with inferior outcomes.1–4
Several mechanisms such as increased state of inflammation, heightened proliferative response to endothelial injury and hypercoagulability are thought to be partly responsible for inferior outcomes in diabetic patients.5 Also, features of the platelets in patients with diabetes are known to be different compared to platelets in subjects without diabetes. Thus, platelets in diabetic patients are known to be larger in size, more adhesive, more prone to activation and to exhibit a greater number of glycoprotein IIb/IIIa receptors.6,7 Therefore, platelet inhibition in relation to ACS is thought to be of even greater importance in diabetes patients.
The introduction of one specific antithrombotic agent, the glycoprotein IIb/IIIa inhibitor (GPI), abciximab, has shown to reduce the risk of ischemic events following PCI in both patients with8–10 and without DM.11–14 Interestingly, not all trials have found this beneficial effect in patients with DM.15 Because of this inconsistency, and because patients in randomized, controlled trials (RCTs) may not always be representative of patients in daily clinical practice,16 we set out to assess the efficacy of abciximab in PCI-treated diabetic patients with ACS in a real-life scenario.
From the aDepartment of Cardiology, Gentofte University Hospital, Denmark, bDepartment of Cardiology, Glostrup University Hospital, Denmark, cDepartment of Medicine, Holbaek Hospital, Denmark, and dClinical Institute of Surgery and Internal Medicine, Faculty of Health Science, University of Copenhagen, Denmark.
The authors report no conflicts of interest regarding the content herein.
Manuscript submitted August 20, 2010, provisional acceptance given August 27, 2010, final version accepted September 21, 2010.
Address for correspondence: Allan Z. Iversen, MD, Department of Cardiology P, Gentofte University Hospital, Niels Andersens Vej 65, DK-2900, Hellerup, Denmark. E-mail: firstname.lastname@example.org
MethodsStudy population. Copenhagen University Hospital Gentofte in Denmark, is a high-volume invasive center with a catchment population of 1.2 million citizens, which is more than 20% of the total Danish population; it is the invasive hub for 10 non-invasive cardiology departments. On-site cardiac surgery is available and > 1,500 PCI procedures are performed annually. Approximately 500 are primary PCI procedures (pPCI), with each individual PCI operator performing approximately 300 PCIs and 100 pPCIs per year. From January 2003 to November 2008, we identified 5,003 consecutive ACS patients (ST-segment elevation myocardial infarction [STEMI], non-ST segment elevation myocardial infarction [NSTEMI] or unstable angina pectoris [UAP]) admitted to our center for PCI. Acute STEMI was defined as the presence of chest pain for > 30 minutes and 4 mm in at least 2 contiguous precordial electrocardiographic (ECG) leads or > 2 mm in at least two or more contiguous limb ECG leads. In these patients, pPCI was performed according to contemporary interventional guidelines using pretreatment with 10,000 IU of unfractionated heparin, which was supplemented if the activated clotting time (ACT) was ®, Eli Lilly, Indianapolis, Indiana) was used as bailout in case of no-reflow, high thrombus burden, dissection or complex lesions at the discretion of the operator. The regimen used was a bolus of 0.25 mg/kg body weight followed by a 12-hour intravenous infusion of 0.125 µg/kg body weight per minute. In the event of more than one PCI procedure within the study period, the first procedure was defined as the index procedure and subsequent procedures as new revascularizations. The study was approved by the Danish Data Protection Agency and complied with the 2nd Declaration of Helsinki. The authors have full access to the data and take responsibility for its integrity. All authors have read and approved the manuscript as written. Grouping. Since statistical analyses showed highly significant interaction between DM and use of abciximab for endpoints (p Baseline and procedural data. The baseline data were prospectively collected from all patients and entered in a dedicated PCI registry. Diabetes was defined as use of antidiabetic medication (oral and/or insulin) on admission. Hypertension was defined as use of blood pressure-lowering drugs on admission. Hypercholesterolemia was defined as use of lipid-lowering drugs on admission. Multivessel disease was defined as 2- or 3-vessel disease and complex lesions were defined as type C lesions. Follow up and study endpoints. The primary study endpoints were all-cause mortality, TVR and myocardial infarction (MI). The combined endpoint was the patient-oriented endpoint: all-cause mortality, TVR or MI as proposed by the Academic Research Consortium (ARC).17 All patients were followed for a maximum of 3 years. Follow-up data on all-cause mortality were collected from The National Person Identification Registry which holds information on vital status (alive, dead or emigrated), while follow-up data on MI and any kind of revascularization initially were collected from the highly validated Danish National Board of Health’s National Patient Registry using ICD-10 codes.18 If an event was registered, it was subsequently validated by cross-checking with hospital source data. Myocardial infarction was defined as recurrent chest pain combined with significant increases in cardiac biomarkers and/or relevant ECG changes. Angiographic data related to revascularizations were validated and supplemented using our local PCI registry. Statistics. Baseline characteristics were compared using the Chi-square test for frequencies and the Student’s unpaired t-test or the Mann-Whitney test for continuous variables. Tests for interaction between DM status and abciximab treatment were performed. Unadjusted Kaplan-Meier plots for the three individual endpoints and the combination of those, stratified by groups according to DM status and abciximab treatment, were compared using the log rank test for all four groups and the DM groups alone. Plots are shown for the full follow-up period, with corresponding numbers at risk and event rates after 1, 2 and 3 years. Hazard ratios (HR) were calculated using Cox proportional hazard regression analyses. In the statistical test, p-values Results Patient population and baseline characteristics (Table 1). A total of 5,003 patients with ACS treated with PCI were identified in our registry. A subgroup of 629 patients had diabetes (12.6%). Two-thirds in the DM group were treated with oral antidiabetic medication, whereas one-third were treated with insulin. Further analysis revealed no interaction between insulin treatment and abciximab. Hence, there was no need for stratification regarding antidiabetic medication. In the Non-DM group, 852 patients (19.5%) received abciximab, whereas the number in the DM group was 169 (26.9%); p Clinical long-term outcome: Full follow up. Unadjusted long-term outcomes are presented in Figure 1. During the full follow-up period with a median time of 864 days (approximately 2.4 years) (interquartile range [IQR]: 742 days), a total of 471 patients died, 245 underwent TVR and 426 experienced MI. All-cause mortality was significantly higher in the DM ÷ ABC group compared to the DM + ABC group (16.8% vs. 8.7%; log rank: p = 0.027). Actually, patients in the DM + ABC group showed an all-cause mortality risk comparable to the Non-DM group (Non-DM + ABC: 10.7%; Non-DM ÷ ABC: 11.8%). For TVR, an even more pronounced effect was observed for DM patients receiving abciximab. The need of TVR was 17.9% in the DM ÷ ABC group, whereas only 5.4% in the DM + ABC group underwent TVR (log rank: p the HR for reaching the combined endpoint in the DM + ABC group to be 1.04, CI was 0.72–1.50, and p = 0.83. The equivalent HR for the DM ÷ ABC group was 2.02, CI was 1.67–2.44, and p reaching the combined endpoint in the DM + ABC group (HR: 0.53; CI: 0.35–0.79; p = 0.002). The same results emerged from the analyses of the individual endpoints: all-cause mortality, TVR and MI. The HR for the DM + ABC group was not significantly different from the reference group (Non-DM ÷ ABC), whereas the HR for the DM ÷ ABC was considerably higher (all-cause mortality, HR: 1.48; CI: 1.11–1.97; p = 0.007; TVR, HR: 4.10; CI: 2.98–5.64; p Clinical short- to intermediate-term outcomes: 30-days and 1 year (Figure 3). In addition to the long-term risk assessment, we performed unadjusted landmark analyses on the DM group after 30 days and 1 year. After 30 days, we found no differences between the groups for any of the endpoints. After 1 year, we found a reduction in the combined endpoint in favor of ABC in the DM group (log rank: p = 0.026). The same was seen for TVR (log rank: p = 0.017), whereas we found no differences for all-cause mortality (log rank: p = 0.19) or MI (log rank: p = 0.44). However, after adjustment for all relevant variables by Cox proportional hazard regression analyses, only risk of TVR after 1 year remained significantly reduced in the DM + ABC group compared to the DM ÷ ABC group (HR: 0.37; CI: 0.17–0.83; p = 0.016), implicating abciximab to independently predict this endpoint. Only a trend toward a reduction in the combined endpoint was observed (HR: 0.69; CI: 0.43–1.05; p = 0.08).
DiscussionIn this study, we set out to evaluate the independent impact of abciximab on clinical outcome in diabetes patients with ACS treated with PCI in a real-life clinical setting. Unadjusted event rates for all-cause mortality, TVR and the combined endpoint (also including MI) were significantly lower in the diabetic patients who received abciximab compared to those who did not. This difference remained significant after adjustment for relevant conventional baseline variables. Actually, we found that by administering abciximab to diabetic patients, their risk of reaching the endpoints was reduced to levels similar to ACS patients without diabetes, even though the diabetic patients had more severe comorbidity and angiographic characteristics — and hence, a worse risk profile. Abciximab was administered more frequently in diabetic patients, which reflects both that the coronary pathology was judged to be more severe in diabetic patients, and that diabetes per se might be considered an indication for abciximab. It should be stressed that among the subgroup of diabetic patients receiving abciximab, we found worse angiographic characteristics, i.e., multivessel disease, longer total stent length and more complex lesions. It is important when interpreting our results that we found a positive effect of abciximab in diabetic patients, even though these patients had an a priori worse prognosis, thus supporting the use of abciximab in DM patients. This additional positive effect of abciximab used as an adjunctive therapy in diabetic patients has been reported before. However, data on this topic are not consistent. The ISAR-SWEET trial showed a reduced need for TVR with abciximab on top of pretreatment with clopidogrel when diabetic patients with CAD underwent PCI.15 However, this reduction did not translate into a reduction in mortality and neither was the risk of reinfarction reduced. Patients enrolled in this study were classified as low-to-intermediate risk and the results may not translate into a setting of high-risk patients, i.e., patients with ACS. These results are in contrast to the findings from a pooled analysis of the EPIC, EPILOG and EPISTENT trials in which Bhatt and colleagues found a reduction in both mortality and reinfarction in diabetic patients treated with abciximab compared to placebo.8 However, the above studies showed event rates not quite similar to our findings. In the pooled analyses by Bhatt et al, the 1-year all-cause mortality rate was 2.5% for the diabetic patients who received abciximab and 4.5% for those who did not, which is less than reported in the ISAR-SWEET trial (4.8% and 5.1%, respectively). We found a higher 1-year all-cause mortality rate for diabetic patients irrespective of treatment with abciximab (DM + ABC: 6.0% and DM ÷ ABC: 9.5%). This trend was also observed for the risk of recurrent MI after 1 year. In the ISAR-SWEET trial, the risk was approximately 4.5% (irrespective of assignment group), whereas Bhatt et al reported a risk reduction from 11.6% to 6.0% when treating diabetic patients with abciximab. Our corresponding numbers were 11.8% and 8.2%, respectively; hence, they are comparable to the results from the pooled analysis. Regarding the need for TVR, there is a large discrepancy between our results and the results from ISAR-SWEET and the work of Bhatt et al. In the latter, no reduction in TVR was observed and the event rate was approximately 25%. In the ISAR-SWEET trial, a reduction from 30% to 23% was seen in the abciximab group. These event rates were much higher than what we observed (11.6% in the DM ÷ ABC group versus 4.5% in the DM + ABC group). However, several aspects in the three studies can account for these differences. First, patients treated within the controlled environment of the RCT do not necessarily resemble the patients treated in a real-life setting. For instance, the mean age at inclusion in the pooled analysis by Bhatt et al was 60 years, whereas the mean age in our study was 67 years. In the ISAR-SWEET trial, only patients with stable ischemic heart disease were included, which probably accounts for the low mortality and risk of MI. Changes in pre- and post-PCI antithrombotic treatment and stent type have occurred since the earliest studies, and the fact that patients were scheduled for repeat angiography in the ISAR-SWEET trial makes it difficult to compare the need for TVR. In a contemporary study, the SORT-OUT II trial, 6.5% of the patients who received DES underwent TVR within the follow-up period of 1.5 years, which is in accordance with our findings.19 Pathophysiological explanations. To understand the mechanisms responsible for the excess risk of ST and restenosis in patients with diabetes, one has to appreciate the pathophysiology of the non-diabetic patient.20 Endothelium of the PCI-treated vessel is exposed to a barotrauma due to balloon inflation and stent deployment. Activated platelets at the culprit site secrete growth factors (GF), which changes the smooth muscle cells (SMC) in the media of the vessel wall from a contractile state to a proliferative and secretory state.21 The SMC then migrate from the media to the intimal part of the vessel wall, thus causing intimal hyperplasia. In patients with diabetes, several mechanisms complicate the response to vessel injury in MI and PCI.22 First, glycosylation and increased production of free oxygen radicals and endothelins, along with impaired prostacyclin production, result in increased platelet aggregation and adhesion, SMC proliferation, inappropriate extracellular matrix deposition, altered GF production and vasoconstriction. In a complex pattern, this leads to decreased reendothelialization, hence increasing the risk of intimal hyperplasia and thrombosis. Second, platelets in patients with diabetes exhibit increased levels of activation, a state in which the glycoprotein IIb/IIIa receptor plays a pivotal role,23 adhesiveness, aggregability and potential for SMC proliferation. Also, increased levels of fibrinogen, factor VII, and decreased fibrinolytic activity due to increased PAI-1 activity lead to increased platelet activation and thrombus formation. With the above mechanisms in mind, it is logical to inactivate platelets (especially in the diabetic patient), e.g., with a GPI such as abciximab. Study limitations. Despite our efforts to reduce bias, the risk of unknown confounders exists in a nonrandomized trial. This is particularly important in a registry study where the treatment-strategy is decided at the operator’s discretion, as in the case with abciximab (and GPIs in general). In the present analyses, our main variable (use of abciximab) could be biased — in this context “confounded by indication.” For example, the PCI operator may have assumed that the overall coronary pathology combined with other comorbidities such as diabetes would favor the use of abciximab. The choice of usage of any GPI is based on relatively complex considerations, which may have been missed in our adjustment. However, if anything, we might have underestimated the effect of abciximab in diabetic patients (and those without diabetes), since the agent is primarily used in patients with a high-risk profile. Furthermore, the prevalence of TVR was also relatively low in our population. Since our study reflects everyday practice, no patients were routinely scheduled for repeat angiography, as is sometimes the case in RCTs. All new revascularizations were symptom-driven and then evaluated thoroughly. Also, we do not have data on bleeding complications, which is a known risk when administering GPIs. However, neither the ISAR-SWEET trial nor the pooled analysis by Bhatt et al showed any significant increased risk of bleeding in diabetic patients receiving abciximab. The low percentage of diabetics in our population (12.6%) could, to some extent, be explained by a lower background prevalence of diabetes in Denmark compared to the United States.24 Identification of diabetic status was made very rigorously in our study and may thus underestimate the actual prevalence. Thus, only patients who were treated with antidiabetic drugs at the time of their index admission were defined as having diabetes. Care should be taken, however, when extrapolating our results to populations with very different distributions of potential risk factors.
ConclusionIn this study we found a marked reduction in all-cause mortality, need for TVR and the composite of all-cause mortality, need for TVR and MI when abciximab was administered to diabetic patients with ACS during PCI. This supports the use of abciximab in diabetic patients in the setting of ACS and PCI.
1. Abizaid A, Kornowski R, Mintz GS, et al. The influence of diabetes mellitus on acute and late clinical outcomes following coronary stent implantation. J Am Coll Cardiol 1998;32:584–589. 2. Barsness GW, Peterson ED, Ohman EM, et al. Relationship between diabetes mellitus and long-term survival after coronary bypass and angioplasty. Circulation 1997;96:2551–2556. 3. Mathew V, Wilson SH, Barsness GW, et al. Comparative outcomes of percutaneous coronary interventions in diabetics vs non-diabetics with prior coronary artery bypass grafting. Eur Heart J 2002;23:1456–1464. 4. Stein B, Weintraub WS, Gebhart SP, et al. Influence of diabetes mellitus on early and late outcome after percutaneous transluminal coronary angioplasty. Circulation 1995;91:979–989. 5. Roffi M, Topol EJ. Percutaneous coronary intervention in diabetic patients with non-ST-segment elevation acute coronary syndromes. Eur Heart J 2004;25:190–198. 6. Tschoepe D, Roesen P, Kaufmann L, et al. Evidence for abnormal platelet glycoprotein expression in diabetes mellitus. Eur J Clin Invest 1990;20:166–170. 7. Tschoepe D, Roesen P, Esser J, et al. Large platelets circulate in an activated state in diabetes mellitus. Semin Thromb Hemost 1991;17:433–438. 8. Bhatt DL, Marso SP, Lincoff AM, et al. Abciximab reduces mortality in diabetics following percutaneous coronary intervention. J Am Coll Cardiol 2000;35:922–928. 9. Lincoff AM, Califf RM, Moliterno DJ, et al. Complementary clinical benefits of coronary-artery stenting and blockade of platelet glycoprotein IIb/IIIa receptors. Evaluation of Platelet IIb/IIIa Inhibition in Stenting Investigators. N Engl J Med 1999;341:319–327. 10. Marso SP, Lincoff AM, Ellis SG, et al. Optimizing the percutaneous interventional outcomes for patients with diabetes mellitus: Results of the EPISTENT (Evaluation of platelet IIb/IIIa inhibitor for stenting trial) diabetic substudy. Circulation 1999;100:2477–2484. 11. The CAPTURE Investigators. Randomised placebo-controlled trial of abciximab before and during coronary intervention in refractory unstable angina: The CAPTURE Study. Lancet 1997;349:1429–1435. 12. Kastrati A, Mehilli J, Neumann FJ, et al. Abciximab in patients with acute coronary syndromes undergoing percutaneous coronary intervention after clopidogrel pretreatment: The ISAR-REACT 2 randomized trial. JAMA 2006;295:1531–1538. 13. Montalescot G, Barragan P, Wittenberg O, et al. Platelet glycoprotein IIb/IIIa inhibition with coronary stenting for acute myocardial infarction. N Engl J Med 2001;344:1895–1903. 14. Topol EJ, Lincoff AM, Kereiakes DJ, et al. Multi-year follow-up of abciximab therapy in three randomized, placebo-controlled trials of percutaneous coronary revascularization. Am J Med 2002;113:1–6. 15. Mehilli J, Kastrati A, Schuhlen H, et al. Randomized clinical trial of abciximab in diabetic patients undergoing elective percutaneous coronary interventions after treatment with a high loading dose of clopidogrel. Circulation 2004;110:3627–3635. 16. Hordijk-Trion M, Lenzen M, Wijns W, et al. Patients enrolled in coronary intervention trials are not representative of patients in clinical practice: Results from the Euro Heart Survey on Coronary Revascularization. Eur Heart J 2006;27:671–678. 17. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials: A case for standardized definitions. Circulation 2007;115:2344–2351.
18. Madsen M, Davidsen M, Rasmussen S, et al. The validity of the diagnosis of acute myocardial infarction in routine statistics: A comparison of mortality and hospital discharge data with the Danish MONICA registry. J Clin Epidemiol 2003;56:124–130. 19. Galloe AM, Thuesen L, Kelbaek H, et al. Comparison of paclitaxel- and sirolimus-eluting stents in everyday clinical practice: The SORT OUT II randomized trial. JAMA 2008;299:409–416. 20. Bauters C, Isner JM. The biology of restenosis. Prog Cardiovasc Dis 1997;40:107–116. 21. Thyberg J, Hedin U, Sjolund M, et al. Regulation of differentiated properties and proliferation of arterial smooth muscle cells. Arteriosclerosis 1990;10:966–990. 22. Aronson D, Bloomgarden Z, Rayfield EJ. Potential mechanisms promoting restenosis in diabetic patients. J Am Coll Cardiol 1996;27:528–535. 23. Panchatcharam M, Miriyala S, Yang F, et al. Enhanced proliferation and migration of vascular smooth muscle cells in response to vascular injury under hyperglycemic conditions is controlled by beta3 integrin signaling. Int J Biochem Cell Biol 2010;42:965–974. 24. Carstensen B, Kristensen JK, Ottosen P, Borch-Johnsen K. The Danish National Diabetes Register: Trends in incidence, prevalence and mortality. Diabetologia 2008;51:2187–2196.