Clinical and Angiographic Outcomes in Diabetic Patients following Single or Multivessel Stenting in the COSTAR II Randomized Tri

The prevalence of diabetes has reached epidemic proportion in many sectors of the world.¹ Atherosclerotic cardiovascular disease has been correlated with the presence, duration and severity of diabetes.^2,3 In addition, both clinical and angiographic outcomes following percutaneous coronary intervention (PCI) are worse in patients with diabetes when compared with their nondiabetic counterparts.^4,5 Although coronary stent implantation improved the outcomes of diabetic patients compared with balloon angioplasty due to a reduction in periprocedural and late (restenosis) complications,^6–8 restenosis following bare-metal stent (BMS) deployment remained a significant limitation to PCI in this patient population.^9,10 Stent-based elution of either paclitaxel or sirolimus from biostable polymers has been demonstrated to reduce angiographic and clinical restenosis compared with BMS deployment in patients both with or without diabetes mellitus.^11–14 However, the safety and efficacy of paclitaxel elution from a stent-based bioresorbable polymer in patients with diabetes mellitus has not been studied. Furthermore, the importance of hyperglycemia in the absence of diagnosed diabetes mellitus as a determinant of clinical and/or angiographic outcomes following coronary stent deployment is inadequately defined.^15,16 Therefore, prespecified subgroup analysis of outcomes in patients with diagnosed diabetes mellitus, and of those with elevated glycolated hemoglobin (HbA1c) in the absence of known diabetes was undertaken from all patients enrolled in the CObalt chromium STent with Antiproliferative for Restenosis (COSTAR) II trial. Study design and patient population. The COSTAR II study design has been described.^17,18 Briefly, COSTAR II is a prospective, multicenter, single-blind, asymmetric (3:2), randomized trial comparing clinical and angiographic outcomes following deployment of the CoStar (Conor MedSystems, Menlo Park, California) compared with the Taxus (Boston Scientific, Natick, Massachusetts) paclitaxel drug-eluting stent (DES) for elective PCI in patients with de novo single- or multivessel coronary artery disease (CAD). The central hypothesis of the COSTAR II trial was that the CoStar stent is noninferior to the Taxus stent for the treatment of patients with symptomatic CAD.
Prespecified subgroup analyses included patients with diabetes, multivessel (versus single-vessel) stented cohorts as well as patients who required the provisional use of overlapping stents. Subject inclusion and exclusion criteria have been described^17,18 and were similar to those of the Taxus IV study,¹² with the exception being the inclusion of patients with 1-, 2- or 3-vessel CAD. Dual antiplatelet therapy (aspirin plus thienopyridine) was prescribed by protocol for 6 months following stent deployment. The study protocol compiled with the Declaration of Helsinki and was approved by the institutional review boards of all participating centers. All participating subjects signed an informed consent.
CoStar stent. The CoStar stent is a novel, thin-strut (0.0035 inch) cobalt chromium alloy metal platform with multiple laser-cut holes within the stent struts that serve as reservoirs for a bioresorbable (poly-lactic-co-glycolic acid; PLGA) polymer matrix. The polymer is loaded with paclitaxel, and the kinetics of both polymer resorption as well as paclitaxel elution are programmable by altering the ratio of copolymer (lactic or glycolic acid) constituents. Based on precedent clinical evaluations, a 10 µg paclitaxel dose eluted over 30 days (in vitro) dose kinetic was used in the present study.¹⁹
Data collection, follow up and core laboratory analyses. All data were submitted to the data coordinating center (Duke Clinical Research Institute) and all clinical events (30 days, 8 and 12 months) were adjudicated by an independent and blinded clinical events committee. All patients underwent baseline angiography. Angiographic (n = 250 multivessel; n = 100 single-vessel PCI patients) and intravascular ultrasound (IVUS) substudies (n = 70 single-vessel PCI patients) were performed. Cardiovascular Research Foundation (New York, New York) served as the angiographic core laboratory and Cardiovascular Core Analysis Lab (Stanford University, Palo Alto, California) served as the IVUS core laboratory. The clinical primary endpoint assessment was completed at 8 months following enrollment, while angiographic endpoints were analyzed 1 month following clinical assessment (at 9 months) to avoid confounding clinical endpoint measures in this single-blind study. An HbA1c level was obtained on all patients at baseline prior to stent deployment.
Endpoint definitions.The primary endpoint of the trial was the aggregate occurrence of MACE out to 8 months including death not attributed to a noncardiac cause or a nonintervention vessel; new Q- or non-Q-wave myocardial infarction (MI) not clearly attributed to a nonintervention vessel (the diagnosis of MI required a CK level > 2 times the upper limit of normal in the presence of elevated CK-MB fraction); and clinically-driven target vessel revascularization (TVR). Key secondary clinical endpoints included device, lesion and procedural success;¹⁷ MACE at 30 days and 12 months post procedure; target lesion revascularization (TLR), clinically-driven TLR and target vessel failure (TVF), defined as the composite of TVR, recurrent MI or cardiac death not attributed to a vessel other than the target vessel. The primary angiographic endpoint of the trial was in-segment late lumen loss with secondary endpoints including in-stent late lumen loss, in-stent and in-segment binary (> 50%) restenosis as well as in-stent and in-segment minimum luminal diameter. An HbA1c level > 6.5% in the absence of previously diagnosed diabetes was felt to be consistent with unrecognized hyperglycemia.²⁰Acute stent thrombosis was defined as abrupt vessel closure of the treatment site resulting in clinical manifestations of ischemia and angiographic evidence of occlusion or flow-limiting thrombosis in a treated vessel in which the investigational device was successfully implanted and that occurred after the procedure, but before the patient left the catheterization laboratory. Subacute stent thrombosis was defined as abrupt vessel closure of the treatment site that resulted in clinical manifestations of ischemia and occlusion occurring after the patient left the catheterization laboratory, but within 30 days of the interventional procedure. Late stent thrombosis was defined as MI attributable to the target vessel, with angiographic documentation of thrombus or total occlusion at the target lesion > 30 days following successful implantation of the device.
Statistical analyses. Prespecified subgroups classified according to presence or absence of diabetes, diabetes treatment and elevated HbA1c levels were studied. Treatment assignments used in these comparisons were based on the intent-to-treat principal.
Continuous data are presented as means and categorical variables are presented as percentages, unless otherwise stated. Selected baseline characteristics, clinical and angiographic outcomes were compared between treatment groups and various diabetes strata by the chi-square test in instances of discrete variables and t-test in instances of continuous variables. Relative risk reduction and 95% confidence intervals (CI) were used for comparisons of major clinical outcomes. A p-value < 0.05 was considered statistically significant. No statistical adjustment was made for multiple comparisons. All statistical analyses were done using SAS version 8.0 or higher.

Results

Between May 23, 2005 and April 20, 2006, 1,700 patients were enrolled at 71 international clinical sites from which 1,675 patients were included in the final analysis. Twenty-five patients were deregistered due to reasons previously described (Krucoff et al, J Am Coll Cardiol, in press). Of all patients enrolled in the trial, 469 had previously diagnosed diabetes (117 insulin-dependent, 351 non-insulin-dependent, 1 unknown) and 77 patients (4.6% of the total study population) had an elevated HbA1c > 6.5% in the absence of previously diagnosed diabetes, and for the purpose of analysis, were considered to have unrecognized hyperglycemia. Thus, 545 (32.5%) patients enrolled in the COSTAR II trial were classified as being hyperglycemic with either clinically recognized DM or not. Baseline clinical demographics and angiographic measures by diabetic status and randomly assigned stent type were similar (Table 1), with the exception that diabetic patients treated with either CoStar or Taxus stents had an increased prevalence of hypertension and hyperlipidemia when compared with their nondiabetic counterparts. Patients with unrecognized hyperglycemia were similar to their nondiabetic counterparts. Compliance with dual antiplatelet therapy (aspirin plus thienopyridine) was high throughout the 6 months prescribed by the study protocol and was not different between randomly assigned stent types (91.7% CoStar, 87.8% Taxus) or by diabetic status (90.8% diabetic, 86.5% nondiabetic). MACE increased in patients with diabetes compared with those without diabetes (12.9 vs. 7.9%, respectively; p = 0.002), and among diabetic patients, MACE appeared highest in those who required insulin compared with those who did not (20.2 vs. 10.6%; p = 0.008). The reduction in MACE attributed to the Taxus stent (24% reduction vs. the CoStar stent) (Figure 1) among patients with diabetes (relative risk ratio for the CoStar stent 1.32 [95% CI 0.80, 2.18]) was less than the relative MACE reduction observed (46%) in the nondiabetic cohort (relative risk ratio for CoStar stent 1.86 [95% CI 1.19, 2.89]). Patients with an elevation in HbA1c > 6.5% in the absence of diagnosed diabetes had a low rate of MACE (2.4% CoStar; 2.8% Taxus). Clinically-driven TVR was increased in diabetics compared with nondiabetics (8.8 vs. 5.7%; p = 0.025) and was highest among diabetics who required insulin compared with those who did not (14.0 vs. 7.1%; p = 0.023). Clinically-driven TVR stratified by stent type and diabetic status is shown in Figure 2. Patients with elevated HbA1c levels in the absence of diagnosed diabetes had low rates of TVR. No statistically significant differences in the incidence or time-course of stent thrombosis through 9 months were observed by randomly assigned stent type (0.6% Costar; 0.1% Taxus) or diabetic status.
Angiographic endpoint analysis. In-segment and in-stent late lumen loss were reduced by Taxus (versus CoStar) stent deployment across all subgroups (Table 2), and in-stent and in-segment binary angiographic restenosis tended to be reduced by Taxus versus CoStar stenting.
IVUS analysis. Analysis of IVUS parameters at 9 months stratified by diabetic status and randomly assigned stent type (Figure 3) demonstrates similar average stent areas. However, neointimal hyperplasia areas and volumes were significantly greater following CoStar deployment compared with Taxus stent deployment. Analysis of glycemic control. Analyses of MACE out to 8 months by quartile or median value for HbA1c (< vs. > 5.8%), as well as by the value of HbA1c used to establish the diagnosis of hyperglycemia (< 6.5% vs. > 6.5%), demonstrated no apparent relationship between baseline glycemic control and late clinical outcomes (Table 3). When analyzed as a continuous function across the entire study population, HbA1c at baseline was correlated with MACE out to 8 months (Odds ratio, OR [95% CI] = 1.16 [1.02, 1.32]; p = 0.020) (Figure 4A). The relationship between HbA1c and MACE was less evident for those patients with diagnosed diabetes (OR [95% CI] 1.10 [0.92,1.31]; p = 0.294) (Figure 4B). Similar analyses of baseline HbA1c correlated with TVR to 8 months for all patients and for those with diagnosed DM (Figures 4 C and D ) demonstrated a weak relationship between baseline glycemic control and revascularization for all subjects (OR [95% CI] = 1.17 [1.01, 1.35]; p = 0.036) which was less apparent for those with diagnosed DM (OR [95% CI] 1.13 [0.93, 1.38]; p = 0.226). No correlation between HbA1c and the occurrence of death or MI was observed either for all patients (OR [95% CI] 1.13 [0.94, 1.36]; p = 0.198) or for those with diagnosed DM (OR [95% CI] 1.06 [0.81, 1.38]; p = 0.681). No significant correlation between baseline HbA1c with subsequent late lumen loss could be discerned for all patients or for those with diagnosed diabetes who underwent follow-up angiography. Multivessel versus single-vessel analysis. Analysis of total MACE and clinically-driven TVR out to 8 months by diabetic status and randomly assigned stent type (Table 4) demonstrates an increased event rate following multivessel (versus single-vessel) stenting and in patients with diabetes (especially insulin requiring) compared with those without diabetes.

Discussion

Previous studies of coronary BMS deployment have demonstrated the independent association of DM with adverse clinical (MACE, TLR, TVR) and angiographic (late lumen loss, binary angiographic restenosis) outcomes.^7–10 Prior diabetic subgroup analyses from the pivotal randomized trials of both the Cypher sirolimus-eluting and the Taxus paclitaxel-eluting DES compared with their respective BMS platforms have demonstrated significant improvement in clinical and angiographic outcomes following DES compared with BMS deployment.13,14
This prespecified diabetic subgroup analysis from the pivotal COSTAR II trial provides the following observations: 1) Following DES deployment, MACE were increased in diabetics (versus nondiabetics) and appears highest in diabetics who require insulin. 2) As a measure of preprocedural glycemic control, HbA1c was weakly correlated with clinical or angiographic outcomes in this trial. 3) MACE (8 months) trended lower and late lumen loss in-segment (9 months) was reduced among diabetic patients treated with Taxus compared to CoStar stent deployment.
Even in the current contemporary era of DES and adjunctive pharmacotherapies, clinical and angiographic outcomes following PCI are worse in patients with diabetes and appear to be directly proportional to the severity of diabetes as reflected by the requirement for insulin treatment. This observation is consistent with prior reports following either bare-metal^7,8,21 or Taxus stent deployment.¹⁴ Indeed, MACE out to 1-year follow up in the Taxus IV trial diabetic substudy were 19.6% for insulin-dependent diabetics compared with 13.6% for diabetics treated with oral medications and 9.4% for nondiabetics.¹⁴ This observation underscores the need for further improvement in diabetic therapies as well as the importance of aggressive medical therapy to achieve recommended glycemic control targets and the management of the usual risk factors in this patient cohort.
The superiority of Taxus compared with CoStar stents in reducing adverse clinical as well as angiographic events in both diabetic and nondiabetic patients may be explained by the higher dose of paclitaxel loaded onto the stent (100 vs. 10 µg) or by differences in polymer drug release kinetics. The relative benefit of Taxus (versus CoStar) is explained by differences in efficacy of restenosis suppression (reduced angiographic and clinical restenosis), as no differences were observed between stents in the incidences of death or MI. Although this study was not adequately powered to detect differences in stent thrombosis, no safety concerns were observed for either stent type, and the incidence of stent thrombosis appeared low and similar between stents. The occurrence of very late stent thrombosis²² cannot be ascertained due to the limited window of observation (12 months) and definitive assessment/comparison of relative thrombosis events between stent platforms awaits longer-term follow up.
Finally, the present analysis suggests that preprocedural HbA1c as a measure of glycemic control is weakly correlated with clinical or angiographic outcomes when evaluated outside of the context of diagnosed diabetes or the type of diabetic treatment. Only by analyzing HbA1c as a continuous function across the entire study population was a statistically significant correlation with MACE or clinically-driven TVR to 8 months evident (Figures 4 A and C). This relationship was less evident for subjects with diagnosed DM. Prior studies have suggested that preprocedural HbA1c > 7.0% is associated with clinical restenosis (increased TVR)¹⁵ and that a fasting blood glucose > 110 mg/dl is associated with an increase in mortality following PCI.¹⁶ Although a correlation between the diagnosis of DM and clinical and angiographic outcomes was observed, the present study provides no measure of either diabetes duration or the type of oral hypoglycemic therapy employed. These factors, which have been correlated with the presence and/or extent of vascular disease, as well as with the frequency occurrence of vascular events in diabetic patients,^23,24 were not prospectively evaluated in this study. Indeed, recent data from randomized, controlled trials suggest that thiazolidinedione therapy reduces both risk of TVR and angiographic coronary late lumen loss.^25,26 Furthermore, although an elevated HbA1c reflects some degree of chronicity in serum glucose elevation, only a single value for HbA1c at baseline was obtained in the CoStar trial. Thus, changes in lifestyle and/or better glycemic control following study enrollment could potentially have contributed to the apparent lack of correlation between baseline HbA1c and subsequent outcomes. Finally, although this analysis was prespecified by protocol, it remains underpowered to provide definitive comparisons between diabetic and nondiabetic cohorts for low-frequency adverse events.

Conclusions

This study demonstrates that patients with diabetes have worse clinical and angiographic outcomes when compared to patients without diabetes following DES deployment. Among diabetic patients, clinical and angiographic outcomes are worse in those who require insulin. The Taxus stent reduced MACE in both diabetic and nondiabetic study cohorts when compared with the CoStar stent. As a measure of preprocedural glycemic control, HbA1c level was weakly correlated with clinical and angiographic outcomes in this contemporary PCI experience. New and better strategies for improving clinical and angiographic outcomes following PCI with DES in diabetic patients are required.

1. Taubert G, Winkelmann BR, Schleiffer T, et al. Prevalence, predictors, and consequences of unrecognized diabetes mellitus in 3266 patients scheduled for coronary angiography. Am Heart J 2003;145:285–291.
2. Lemp GF, Vander Zwaag R, Hughes JP, et al. Association between the severity of diabetes mellitus and coronary arterial atherosclerosis. Am J Cardiol 1987;60:1015–1019.
   
3. Beckman J, Creager MA, Libby P. Diabetes and atherosclerosis: Epidemiology, pathophysiology, and management. JAMA 2002;287:2570–2581.
   
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. Niles NW, McGrath PD, Malenka D, et al. Survival of patients with diabetes and multivessel coronary artery disease after surgical or percutaneous coronary revascularization; Results of a large regional prospective study: Northern New England Cardiovascular Disease Study Group. J Am Coll Cardiol 2001;37:1008–1015.
   
6. Savage MP, Fischman DL, Schatz RA, et al. Coronary intervention in the diabetic patient; Improved outcome following stent implantation compared with balloon angioplasty. Clinical Cardiology 2002;25:213–217.
   
7. 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.
   
8. Elezi S, Kastrati A, Pache J, et al. Diabetes mellitus and the clinical and angiographic outcome after coronary stent placement. J Am Coll Cardiol 1998;32:1866–1873.
   
9. Van Belle E, Bauters C, Hubert E, Bodart JC, et al. Restenosis rates in diabetic patients: a comparison of coronary stenting and balloon angioplasty in native coronary vessels. Circulation 1997;96:1454–1460.
   
10. Van Belle E, Périé M, Braune D, et al. Effects of coronary stenting on vessel patency and long-term clinical outcome after percutaneous coronary revascularization in diabetic patients. J Am Coll Cardiol 2002;40:410–417.
    
11. Moses JW, Leon MB, Popma JJ, et al. SIRIUS Investigators. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315–1323.
    
12. Stone GW, Ellis SG, Cox DA, et al. TAXUS-IV Investigators. A polymer-based, paclitaxel-eluting stent in patients with coronary artery disease. N Engl J Med 2004;350:221–231.
    
13. Moussa I, Leon MB, Baim DS, et al. Impact of sirolimus-eluting stents on outcome in diabetic patients – A SIRIUS (SIRolImUS-coated Bx Velocity balloon-expandable stent in the treatment of patients with de novo coronary artery lesions) substudy. Circulation 2004;109:2273–2278.
    
14. Hermiller JB, Raizner A, Cannon L, et al.; TAXUS-IV Investigators. Outcomes with the polymer-based paclitaxel-eluting TAXUS stent in patients with diabetes mellitus — The Taxus IV trial. J Am Coll Cardiol 2005;45:1172–1179.
    
15. Corpus RA, George PB, House JA, et al. Optimal glycemic control is associated with a lower rate of target vessel revascularization in treated type II diabetic patients undergoing elective percutaneous coronary intervention. J Am Coll Cardiol 2004;43:8–14.
    
16. Muhlestein JB, Anderson JL, Horne BD, et al.; Intermountain Heart Collaborative Study Group. Effect of fasting glucose levels on mortality rate in patients with and without diabetes mellitus and coronary artery disease undergoing percutaneous coronary intervention. Am Heart J 2003;146:351–358.
    
17. Wang TY, Hasselblad V, Peterson JL, et al. The Cobalt chromium STent with Antiproliferative for Restenosis II (COSTAR II) trial study design: Advancing the active-control evaluation of second-generation drug-eluting stents. Am Heart J 2007;153:743–748,
    
18. Krucoff MW, Kereiakes DJ, Petersen JL, et al. Evaluation of a novel bioresorbable polymer paclitaxel eluting stent for the treatment of single and multivessel coronary disease: primary results of the Cobalt Chromium Stent with Antiproliferative for Restenosis II (COSTAR II) Study. J Am Coll Cardiol (in press).
    
19. Dawkins K, Verheye S, Shuhlen H, et al. The European cobalt STent with Antiproliferative for Restenosis trial. (EuroSTAR): 12 month results. EuroIntervention 2007;3:82–88.
    
20. Corbin R, Davidson J, Ganda O, et al. American College of Endocrinology consensus statement on guidelines for glycemic control. Endocr Pract 2002;8(Suppl 1):5–11.
    
21. Schofer J, Schluter M, Rau T, et al. Influence of treatment modality on angiographic outcome after coronary stenting in diabetic patients: A controlled study. J Am Coll Cardiol 2000;35:1554–1559.
    
22. Cutlip DE, Windecker S, Mehran R, et al. Academic Research Consortium. Clinical end points in coronary stent trials: A case for standardized definitions. Circulation 2007;115:2344–2351.
    
23. McGuire DK, Newby CK, Bhapkar MV, et al; SYMPHONY and 2nd SYMPHONY Investigators. Association of diabetes mellitus and glycemic control strategies with clinical outcomes after acute coronary syndromes. Am Heart J 2004;147:246–252.
    
24. Kao J, Tobis J, McClelland RT, et al, Investigators in the Prevention of Restenosis With Tranilast and Its Outcomes Trial. Relation of metformin treatment to clinical events in diabetic patients undergoing percutaneous intervention. Am J Cardiol 2004;93:1347–1350.
    
25. Riche DM, Valderrama R, Henyan NN. Thiazolidinediones and risk of repeat target vessel revascularization following percutaneous coronary intervention. Diabetes Care 2007;30:384–388.
    
26. Marx N, Wöhrle J, Nusser T, et al. Pioglitazone reduces neointima volume after coronary stent implantation: A randomized, placebo-controlled, double blind trial in nondiabetic patients. Circulation 2005;112:2792–2798.