Conflict of interest: Nothing to report.
1522-726X/asset/olbannerleft.gif?v=1&s=168f8bf3d7f4750c78fb810aa2fa1336c10e443a)
1522-726X/asset/olbannerright.gif?v=1&s=741c38da03ec7f2645d5f0ce328e5c02cd630303)
Coronary Artery Disease
You have free access to this content
Paclitaxel versus sirolimus eluting stents in diabetic patients: Does stent type and/or stent diameter matter?: Long-term clinical outcome of 2,429-patient multicenter registry†
Article first published online: 8 NOV 2012
DOI: 10.1002/ccd.24445
Copyright © 2012 Wiley Periodicals, Inc.
Issue
Catheterization and Cardiovascular Interventions
Volume 81, Issue 1, pages 80–89, 1 January 2013
Additional Information
How to Cite
Tarantini, G., Facchin, M., Capodanno, D., Musumeci, G., Saia, F., Menozzi, A., Meliga, E., Mancone, M., Lettieri, C. and Tamburino, C. (2013), Paclitaxel versus sirolimus eluting stents in diabetic patients: Does stent type and/or stent diameter matter?: Long-term clinical outcome of 2,429-patient multicenter registry. Cathet. Cardiovasc. Intervent., 81: 80–89. doi: 10.1002/ccd.24445
- †
Publication History
- Issue published online: 21 DEC 2012
- Article first published online: 8 NOV 2012
- Accepted manuscript online: 18 APR 2012 04:54AM EST
- Manuscript Accepted: 12 APR 2012
- Manuscript Received: 7 OCT 2011
Keywords:
- diabetes;
- coronary artery disease;
- stent;
- percutaneous coronary intervention
Abstract
Background
Drug-eluting stents (DES) are more effective in reducing restenosis than bare-metal stents. Less certain is the relative performance of the two widely used DES—sirolimus-eluting stents (SES) and paclitaxel-eluting stents (PES)—in diabetic patients undergoing percutaneous coronary intervention (PCI).
Objectives
We studied the long-term effectiveness and safety of SES versus PES in diabetic patients, overall and grouped by the size of the stent placed in the native-stented vessel.
Methods
Data were obtained from an observational, multicenter registry of 2,429 consecutive patients with DM who underwent PCI between August 2003 and August 2009 with SES (n = 1,370) or PES (n = 1,059). Data were analyzed separately for patients with small stents (<3.0 mm, n = 1,274) vs. only large (≥3.0 mm, n = 1,155).
Results
At Cox-adjusted analysis of the overall cohort, there was no significant difference between SES and PES with regard to death/myocardial infarction (D/MI) (P = 0.6) or target lesion revascularization (TLR) (P = 0.3) either in “small-stent” (D/MI, P = 0.8; TLR, P = 0.2) or “large stent” group (D/MI, P = 0.8; TLR, P = 0.4) throughout 1 to 5-year follow-up. These results were confirmed by sensitivity, propensity-score matched analysis (717 matched pairs) that failed to find differences in the safety and efficacy between SES and PES.
Conclusions
In this large observational study, PES and SES were equally safe and efficacious in diabetic patients undergoing PCI in clinical practice, regardless of the stent size. © 2012 Wiley Periodicals, Inc.
INTRODUCTION
Compared with bare metal stents (BMS), the use of drug eluting stents (DES) in diabetic patients is associated with a lower risk for target lesion revascularization (TLR) without any increase in death or myocardial infarction (MI) [1, 2]. Nevertheless, there is some uncertainty about the relative effectiveness and safety of sirolimus-eluting stents (SES) versus paclitaxel-eluting stents (PES) in coronary artery disease patients with diabetes mellitus (DM). Direct head-to-head comparisons outlined that SES are associated with lower late luminal loss compared with PES [3, 4]. However, although late luminal loss has been proposed as a robust marker for discriminating DES in the overall population [5], its impact on clinical outcomes in DM patients, such as TLR, remains controversial, particularly in the low range of late loss typical for DES. Three randomized controlled trials have directly compared the two DES in diabetic patients [6–8]. A major limitation of these studies has been a lack of adequate sample sizes to show the superiority of one DES over the other, the very selected DM population, the short follow-up, and the primarily angiographic, not clinical end points. Adding to the paucity of data on the impact of different 1st generation DES on clinical outcomes of DM patients, our understanding is puzzled by the notion that outcomes associated with each stent may vary depending on the caliber of the stent as surrogate of vessel size [9, 10]. On this background, we studied the long-term safety and efficacy of PES versus SES in a large, real-life, population of DM patients treated by percutaneous coronary intervention (PCI) throughout 1 to 5-year of follow-up stratified by the size of the stent used in the native vessel.
METHODS
Settings, Data Collection, and Interventions
Data were obtained from a retrospective, observational, multicenter registry in which nine high-volume participating centers enrolled 2,429 consecutive patients with DM who underwent PCI between August 2003 and August 2009 with first generation DES (SES [Cypher, Cordis, Johnson & Johnson, Warren, N.J.] or PES [Taxus, Boston Scientific, Natick, Massachusetts]) and eligible at ≥1-year follow-up. Details on setting, data collection intervention, and follow-up are reported in the Supporting Information appendix.
Patient's Subgroups and End-Points
DM patients were grouped by the DES type (PES versus SES) and by the size of the stent (“small-stent” versus “large-stent”) as also reported by others [11]. Patients with at least one small stent (<3.0 mm stents) implanted were categorized a priori as “small-stent” patients and those with only large (≥3.0 mm stents) native vessel stenting as “large-stent” patients. The use of at least one small stent (<3.0 mm diameter) showed a significant impact on target lesion and vessel revascularization in multivariable analyses of our study population. Diabetic patients were categorized as insulin or non-insulin treated. Major adverse cardiac events (MACEs) were defined as a composite of death, MI, or TLR. Death was defined as mortality from all causes. Non-Q wave and Q wave MI were defined as the elevation of the serum creatinine kinase isoenzyme MB or troponin at least three times the upper limit of normal in the absence or presence of new pathological Q waves. TLR was defined as any repeat percutaneous intervention of the target lesion or bypass surgery of the target vessel performed for restenosis or other complication of the target lesion. Target vessel revascularization (TVR) was defined as any repeat percutaneous intervention or surgical bypass of any segment of the target vessel. All the repeated coronary interventions were clinically driven.
Statistical Analysis
Data were expressed as mean and standard deviation or median and interquartile range (IQR) for continuous variables and as counts and percentages for categorical variables. Clinical, angiographic, and procedural data were compared between groups by Student t test or chi-square test, as appropriate. Additional details on statistical methods are described in the Supporting Information appendix. Briefly, Cox Proportional Hazard analysis was conducted to compare the groups defined above with the clinical events (MACE, death or myocardial infarction, TLR, and TVR). As a sensitivity analysis, we conducted a propensity analysis as reported by others [12–14]. The results are expressed as hazard-ratio and relative 95% confidence interval and P values were considered statistically significant when P < 0.05. Data were analyzed with SAS 9.1.3 for Windows, SAS Institute Inc., Cary, NC.
RESULTS
Characteristics of Patients and Procedures
The flow diagram of the study population subdivided according to the DES type and stent size before and after propensity score matching is shown in Fig. 1. Baseline clinical and angiographic characteristics of the whole population as well as of the two subgroups are summarized in Table I. Overall, patients represent a typical “real life” population with DM with high rate of cardiovascular risk factors, previous MI and revascularization rate, and multivessel disease with several high risk characteristics, such as number of the diseased lesions, multiple, and long vessel stenting. The DES use was off-label in 1,579 patients (65%) without difference between stent type and stent group. Baseline characteristics differed significantly between SES and PES either in total population or subgroups (Table I). Matching on propensity score resulted in the formation of the 717 pairs of patients in the overall cohort. All baseline characteristics of the patients, including the mean number of days follow up were well balanced between the PES and SES groups (standardized differences <0.09). All variables were also well balanced among the 357 matched pairs of “large-stent” patients (standardized differences <0.09) and among the 360 matched pairs of “small-stent” patients (standardized differences <0.08) (Table II).
Figure 1. Patients' flow diagram.
| Large stent | Small stent | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SES 1,370 | PES 1,059 | P- value | Large 1,155 | Small 1,274 | P- value | SES 604 | PES 551 | P- value | SES 766 | PES 508 | P- value | ||
| |||||||||||||
| Months of follow-up, median (Q1–Q3) | 25 (12-41) | 23 (12-38) | 0.002 | 23 (12-38) | 25 (12-41) | 0.024 | 24 (12-38) | 23 (12-38) | 0.245 | 26 (13-43) | 23 (12-37) | 0.001 | |
| Male, n (%) | 979 (72) | 761 (72) | 0.846 | 882 (77) | 858 (68) | <0.0001 | 464 (77) | 418 (76) | 0.630 | 515 (68) | 343 (68) | 0.874 | |
| Age (mean ± sd) | 66 ± 10 | 67 ± 9 | 0.021 | 66 ± 10 | 66 ± 10 | 0.675 | 66 ± 10 | 67 ± 9 | 0.188 | 66 ± 10 | 67 ± 9 | 0.057 | |
| Diabetes mellitus | 0.332 | 0.198 | 0.095 | 0.920 | |||||||||
| Insulin requiring DM, n (%) | 445 (33) | 364 (35) | 370 (32) | 439 (35) | 180 (30) | 190 (35) | 265 (35) | 174 (34) | |||||
| Non-insulin requiring DM, n(%) | 919 (67) | 691 (65) | 781 (68) | 829 (65) | 421 (70) | 360 (65) | 498 (65) | 331 (65) | |||||
| Smoke (%) | 0.420 | 0.806 | 0.152 | 0.945 | |||||||||
| Current, n (%) | 185 (14) | 151 (14) | 155 (13) | 181 (14) | 78 (13) | 77 (14) | 107 (14) | 74 (15) | |||||
| Previous, n (%) | 374 (27) | 309 (29) | 330 (29) | 353 (28) | 160 (27) | 170 (31) | 214 (28) | 139 (28) | |||||
| Hypertension, n (%) | 1,110 (81) | 973 (84) | 0.046 | 936 (81) | 1,057 (83) | 0.170 | 477 (79) | 557 (839 | 0.097 | 628 (82) | 432 (85) | 0.185 | |
| Hypercholesterolemia, n (%) | 918 (67) | 720 (68) | 0.794 | 774 (67) | 854 (67) | 0.973 | 423 (70) | 353 (64) | 0.043 | 498 (65) | 361 (71) | 0.020 | |
| Family history, n (%) | 521 (38) | 424 (40) | 0.203 | 404 (35) | 535 (42) | 0.002 | 211 (35) | 198 (36) | 0.557 | 306 (40) | 229 (45) | 0.098 | |
| Renal failure requiring dialysis, n (%) | 28 (2.0) | 25 (2.4) | 0.821 | 26 (2.3) | 27 (2.1) | 0.546 | 14 (2.3) | 12 (2.2) | 0.912 | 14 (2.4) | 13 (2.5) | 0.742 | |
| Primary PCI, n (%) | 178 (13) | 116 (10) | 0.074 | 139 (12) | 140 (11) | 0.389 | 85 (14) | 61 (11) | 0.127 | 92 (12) | 51 (10) | 0.270 | |
| Previous AMI | 0.026 | 0.854 | 0.408 | 0.021 | |||||||||
| Q type, n (%) | 466 (34) | 360 (34) | 393 (34) | 420 (33) | 199 (33) | 198 (36) | 260 (34) | 163 (32) | |||||
| Non-Q type, n (%) | 41 (3) | 11 (1) | 35 (3) | 38 (3) | 18 (3) | 11 (2) | 31 (4) | 5 (1) | |||||
| Previous PCI, n (%) | 438 (32) | 349 (33) | 0.596 | 370 (32) | 420 (33) | 0.564 | 193 (32) | 1,7632 | 0.835 | 245 (32) | 173 (34) | 0.546 | |
| Angina, n (%) | 1,000 (76) | 858 (81) | 0.031 | 901 (78) | 1,006 (79) | 0.662 | 453 (75) | 451 (82) | 0.015 | 597 (78) | 406 (80) | 0.473 | |
| EF% (mean ± sd) | 52 ± 13 | 51 ± 11 | 0.032 | 52 ± 13 | 52 ± 12 | 0.835 | 523 ± 13 | 51 ± 12 | 0.060 | 52 ± 12 | 51 ± 11 | 0.239 | |
| Multivessel disease, n (%) | 949 (70) | 741 (70) | 0.774 | 774 (67) | 916 (72) | 0.008 | 399 (67) | 375 (68) | 0.511 | 550 (72) | 366 (72) | 0.967 | |
| Unknown 0.3% | |||||||||||||
| No. of vessel disease | 0.245 | 0.653 | 0.279 | 0.569 | |||||||||
| 1 vessel, n (%) | 413 (30) | 313 (30) | 375 (33) | 351 (28) | 201 (34) | 174 (32) | 212 (28) | 139 (28) | |||||
| 2 vessels, n (%) | 458 (34) | 404 (38) | 390 (34) | 472 (37) | 191 (32) | 199 (36) | 267 (35) | 205 (41) | |||||
| 3 vessels, n (%) | 488 (36) | 336 (32) | 382 (33) | 442 (35) | 207 (35) | 175 (32) | 281 (37) | 161 (32) | |||||
| Calcium, n (%) | 157 (12) | 123 (12) | 0.985 | 127 (12) | 153 (13) | 0.469 | 55 (10) | 72 (14) | 0.055 | 102 (14) | 51 (11) | 0.092 | |
| Unknown 4.5% | |||||||||||||
| Restenosis, n (%) | 91 (7) | 49 (5) | 0.034 | 72 (6) | 68 (5) | 0.348 | 46 (8) | 26 (5) | 0.041 | 45 (6) | 23 (5) | 0.598 | |
| Ostial, n (%) | 103 (8) | 108 (10) | 0.020 | 143 (12) | 68 (5) | <0.0001 | 60 (10) | 83 (15) | 0.009 | 43 (6) | 25 (5) | 0.59 | |
| Unknown 0.3% | |||||||||||||
| Bifurcation, n (%) | 193 (14) | 176 (17) | 0.085 | 178 (16) | 191 (15) | 0.769 | 80 (13) | 98 (18) | 0.034 | 113 (15) | 78 (15) | 0.753 | |
| Unknown 0.3% | |||||||||||||
| Total occlusion, n (%) | 100 (7) | 63 (6) | 0.186 | 70 (6) | 93 (7) | 0.220 | 36 (6) | 34 (6) | 0.892 | 64 (8) | 29 (6) | 0.077 | |
| Stent length (mean ± sd) | 20 ± 6 | 19 ± 6 | 0.001 | 20 ± 6 | 20 ± 6 | 0.781 | 20 ± 6 | 20 ± 6 | 0.687 | 20 ± 6 | 19 ± 5 | <0.0001 | |
| Stent diameter (mean ± sd) | 2.8 ± 0.4 | 2.9 ± 0.5 | 0.001 | 3.1 ± 0.4 | 2.6 ± 0.4 | <0.0001 | 3.1 ± 0.3 | 3.1 ± 0.6 | 0.024 | 2.6 ± 0.4 | 2.6 ± 0.2 | 0.883 | |
| No lesions (mean ± sd) | 2.6 ± 1.6 | 2.6 ± 1.6 | 0.714 | 2.4 ± 1.5 | 2.8 ± 1.7 | <0.0001 | 2.4 ± 1.4 | 2.5 ± 1.5 | 0.534 | 2.8 ± 1.8 | 2.8 ± 1.7 | 0.643 | |
| No lesions treated (mean ± sd) | 1.5 ± 0.9 | 1.6 ± 0.9 | 0.089 | 1.4 ± 0.8 | 1.6 ± 1.0 | <0.0001 | 1.4 ± 0.8 | 1.5 ± 0.9 | 0.063 | 1.6 ± 0.9 | 1.7 ± 1.0 | 0.233 | |
| No stent per patient (mean ± sd) | 1.8 ± 1.1 | 1.9 ± 1.2 | 0.044 | 1.7 ± 1.0 | 2.1 ± 1.3 | <0.0001 | 1.6 ± 0.9 | 1.8 ± 1.1 | 0.013 | 2.0 ± 1.2 | 2.2 ± 1.4 | 0.125 | |
| No stent per lesion (mean ± sd) | 1.2 ± 0.5 | 1.3 ± 0.5 | 0.519 | 1.2 ± 0.4 | 1.3 ± 0.6 | <0.0001 | 1.2 ± 0.4 | 1.2 ± 0.4 | 0.748 | 1.3 ± 0.6 | 1.3 ± 0.6 | 0.190 | |
| Large stent | Small stent | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SES 717 | PES 717 | Stand. diff. | SES 358 | PES 358 | Stand. diff. | SES 359 | PES 359 | Stand. diff. | |
| |||||||||
| Months of follow-up, median (Q1-Q3) | 24.3 (12-39) | 24.2 (13-39) | 0.011 | 24.0 (12-38) | 24.3 (13-40) | 0.020 | 24.5 (12-39) | 24.2 (13-38) | 0.002 |
| Male, n (%) | 520 (73) | 517 (72) | 0.009 | 273 (76) | 274 (77) | 0.007 | 247 (69) | 243 (68) | 0.024 |
| Age (mean ± sd) | 66.1 ± 9.6 | 66.0 ± 8.9 | 0.003 | 65.5 ± 9.7 | 65.8 ± 9.0 | 0.025 | 66.6 ± 9.3 | 66.3 ± 8.9 | 0.033 |
| Diabetes mellitus | |||||||||
| Insulin Requiring DM, n (%) | 231 (32) | 235 (33) | 0.012 | 104 (29) | 108 (30) | 0.025 | 127 (35) | 127 (35) | <0.001 |
| Non Insulin requiring DM, n (%) | 486 (68) | 482 (67) | 0.012 | 253 (71) | 249 (70) | 0.025 | 233 (65) | 233 (65) | <0.001 |
| Smoke | |||||||||
| Never, n (%) | 384 (54) | 390 (54) | 0.017 | 191 (54) | 190 (53) | 0.006 | 193 (54) | 200 (56) | 0.039 |
| Previous, n (%) | 219 (31) | 208 (29) | 0.034 | 110 (31) | 110 (31) | <0.001 | 109 (30) | 98 (27) | 0.068 |
| Current, n (%) | 114 (16) | 119 (17) | 0.019 | 56 (16) | 57 (16) | 0.008 | 58 (16) | 62 (17) | 0.030 |
| Hypertension, n (%) | 580 (81) | 584 (81) | 0.014 | 275 (77) | 286 (80) | 0.075 | 305 (85) | 298 (83) | 0.053 |
| Hypercholesterolemia, n ( %) | 462 (64) | 460 (64) | 0.006 | 236 (66) | 219 (61) | 0.099 | 226 (63) | 241 (67) | 0.087 |
| Family history, n ( %) | 269 (38) | 288 (40) | 0.054 | 119 (33) | 132 (37) | 0.076 | 150 (42) | 156 (43) | 0.034 |
| Primary PCI, n (%) | 65 (9) | 65 (9) | <0.001 | 40 (11) | 36 (10) | 0.036 | 25 (7) | 29 (8) | 0.042 |
| Previous AMI | |||||||||
| No, n (%) | 473 (66) | 466 (65) | 0.021 | 233 (65) | 224 (63) | 0.053 | 240 (67) | 242 (67) | 0.012 |
| Q type, n (%) | 236 (33) | 243 (34) | 0.021 | 120 (34) | 129 (36) | 0.053 | 116 (32) | 114 (32) | 0.012 |
| Non Q type, n (%) | 8 (1) | 8 (1) | <0.001 | 4 (1) | 4 (1) | <0.001 | 4 (1) | 4 (1) | 0.000 |
| Previous revascularization, n (%) | 216 (30) | 212 (30) | 0.012 | 101 (28) | 98 (27) | 0.019 | 115 (32) | 114 (32) | 0.006 |
| EF% (mean ± sd) | 51.8 ± 13.3 | 51.7 ± 11.4 | 0.007 | 52.2 ± 13.4 | 52.3 ± 11.5 | 0.011 | 51.3 ± 13.1 | 50.0 ± 11.2 | 0.025 |
| Multivessel disease, n (%) | 511 (71) | 517 (72) | 0.019 | 244 (68) | 249 (70) | 0.045 | 267 (74) | 268 (74) | 0.006 |
| No of vessel disease | |||||||||
| 1 vessel, n (%) | 206 (29) | 199 (28) | 0.006 | 113 (32) | 108 (30) | 0.030 | 93 (26) | 91 (25) | 0.019 |
| 2 vessels, n (%) | 251 (35) | 267 (37) | 0.070 | 122 (34) | 127 (36) | 0.035 | 129 (36) | 140 (39) | 0.104 |
| 3 vessels, n (%) | 260 (36) | 251 (35) | 0.064 | 122 (34) | 122 (34) | 0.006 | 138 (38) | 129 (36) | 0.121 |
| Calcium, n (%) | 90 (13) | 87 (12) | 0.013 | 39 (11) | 47 (13) | 0.069 | 51 (14) | 40 (11) | 0.092 |
| Restenosis, n (%) | 35 (5) | 33 (5) | 0.013 | 21 (6) | 17 (5) | 0.050 | 14 (4) | 16 (4) | 0.028 |
| Ostial, n (%) | 66 (9) | 67 (9) | 0.005 | 39 (11) | 49 (14) | 0.085 | 27 (8) | 18 (5) | 0.072 |
| Bifurcation, n (%) | 128 (18) | 121 (17) | 0.026 | 65 (18) | 63 (18) | 0.015 | 63 (186) | 58 (16) | 0.037 |
| Total occlusion, n (%) | 46 (6) | 41 (6) | 0.029 | 18 (5) | 22 (6) | 0.049 | 28 (8) | 19 (5) | 0.070 |
| Stent lenght (mean ± sd) | 19.8 ± 6.2 | 19.8 ± 5.9 | 0.009 | 19.8 ± 6.3 | 20.4 ± 6.4 | 0.072 | 19.9 ± 6.1 | 19.1 ± 5.3 | 0.093 |
| Stent diameter (mean ± sd) | 2.8 ± 0.3 | 2.9 ± 0.6 | 0.087 | 3.1 ± 0.3 | 3.1 ± 0.7 | 0.107 | 2.6 ± 0.2 | 2.6 ± 0.2 | 0.012 |
| No lesions (mean ± sd) | 2.7 ± 1.6 | 2.8 ± 1.7 | 0.035 | 2.5 ± 1.5 | 2.6 ± 1.7 | 0.068 | 2.9 ± 1.7 | 2.9 ± 1.8 | 0.006 |
| No lesions treated (mean ± sd) | 1.6 ± 0.9 | 1.6 ± 0.9 | 0.027 | 1.4 ± 0.8 | 1.5 ± 0.9 | 0.096 | 1.7 ± 1.0 | 1.7 ± 1.0 | 0.031 |
| No stent per patient (mean ± sd) | 1.9 ± 1.2 | 2.0 ± 1.3 | 0.047 | 1.6 ± 1.0 | 1.8 ± 1.1 | 0.074 | 2.2 ± 1.4 | 2.2 ± 1.4 | 0.004 |
| No stent per lesion (mean ± sd) | 1.2 ± 0.5 | 1.3 ± 0.5 | 0.046 | 1.2 ± 0.4 | 1.2 ± 0.4 | 0.013 | 1.3 ± 0.6 | 1.3 ± 0.6 | 0.072 |
Clinical Outcomes
During the entire study period with a median follow-up of 25 months (range 1–5 years) a cumulative total 642 events occurred in our study population. The cumulative incidences of the MACE, D/MI, TLR, and TVR in the total crude population as well as in the different subgroups of patients are shown in Fig. 2. The incidence of TLR increased progressively along each year for both SES and PES use. The unadjusted analysis revealed that the overall MACE rate did not differ significantly between the SES and PES groups. There was a trend toward higher MACEs rate in the “small-stent” group compared with the “large stent” group that increased progressively after 24 months of follow-up, mainly driven by the increase in the TLR and TVR rates (Fig. 2). The D/MI rate was not statistically different among groups. The TLR rate was significantly higher in the “small-stent” group compared with the “large stent” group, and in the patients treated by PES instead of SES in the “small-stent” group. Similarly, the TVR rate was significantly higher in the “small-stent” group compared with the “large stent” group with an unfavorable trend in the patients treated by PES in the “small-stent” group.
Figure 2. Unadjusted incidence of MACE, death/ MI, TLR, and TVR in the total study population and in the subgroups. MI, myocardial infarction; TLR, target lesion revascularization; TVR, target vessel revascularization. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
Table III shows the results of the multivariate Cox-adjusted analysis of the unmatched sample of 2,419 patients. Throughout the follow-up period, the “small-stent” group trended toward a higher incidence of MACE and an higher rate of TLR and TVR compared with the “large stent” group. On the contrary, there were no statistically significant differences between SES and PES in the incidence of MACE, D/MI, TLR. Nevertheless, the patients treated by PES in the “small-stent” group had a lower rate of TVR. When subjects were stratified for insulin treatment, there were no significant differences between PES and SES in the incidence of MACE, D/MI, TLR overall, with only a trend towards lower MACES in non-insulin dependent DM patients treated by PES in the “small-stent” group (Table 4 in Supporting Information appendix).
| PES/SES | Small/large | Large (PES/SES) | Small (PES/SES) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable | HR | 95% CI | P | HR | 95%CI | P | HR | 95% CI | P | HR | 95% CI | P |
| ||||||||||||
| MACE | ||||||||||||
| Unadjusted | 1.03 | 0.85–1.24 | 0.79 | 1.08 | 0.92–1.26 | 0.35 | 1.10 | 0.83–1.45 | 0.51 | 1.00 | 0.78–1.29 | 0.98 |
| Adjusted | 0.84 | 0.65–1.10 | 0.21 | 1.23 | 0.97–1.56 | 0.083 | 1.06 | 0.79–1.43 | 0.69 | 0.83 | 0.63–1.01 | 0.17 |
| MI/DEATH | ||||||||||||
| Unadjusted | 0.93 | 0.72–1.20 | 0.59 | 0.90 | 0.70–1.16 | 0.42 | 0.89 | 0.62–1.28 | 0.53 | 0.96 | 0.66–1.38 | 0.81 |
| Adjusted | 0.91 | 0.62–1.34 | 0.65 | 1.04 | 0.76–1.43 | 0.82 | 0.97 | 0.66–1.41 | 0.86 | 0.94 | 0.64–1.40 | 0.94 |
| TLR | ||||||||||||
| Unadjusted | 1.17 | 0.89–1.53 | 0.26 | 1.49 | 1.13–1.97 | .005 | 1.61 | 1.02–2.53 | 0.040 | 1.05 | 0.74–1.49 | 0.78 |
| Adjusted | 0.80 | 0.56–1.16 | 0.24 | 1.45 | 1.01–2.10 | 0.047 | 1.25 | 0.77–2.12 | 0.36 | 0.75 | 0.52–1.10 | 0.14 |
| TVR | ||||||||||||
| Unadjusted | 1.03 | 0.83–1.29 | 0.78 | 1.33 | 1.06–1.67 | 0.014 | 1.44 | 1.01–2.06 | 0.045 | 0.88 | 0.65–1.18 | 0.39 |
| Adjusted | 0.79 | 0.58–1.07 | 0.12 | 1.36 | 1.01–1.82 | 0.043 | 1.17 | 0.80–1.71 | 0.42 | 0.68 | 0.49–0.94 | 0.018 |
Figure 3 shows the cumulative incidences of MACE, D/MI, TLR, TVR of the 717 matched pair of diabetic patients treated by SES or PES in the overall population as well as in “small-stent” or the “large stent” group. The results of this sensitivity analysis were generally consistent with the Cox analysis of the unmatched cohort, with the exception in the results of TVR that was not different between SES and PES in the small stent group. Similarly to the crude population, also in the propensity-matched population the TLR increased progressively along each year both in SES and PES group.
Figure 3. Incidence of MACE, death/ MI, TLR, and TVR after propensity score matched in the total study population and in the subgroups. MI, myocardial infarction; TLR, target lesion revascularization; TVR, target vessel revascularization. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
DISCUSSION
This is among the largest clinical studies that analyze the efficacy and safety of the choice of the SES instead of the PES in an unselected, real life, cohort of diabetic patients throughout 1 to 5-year of follow-up also as function of stent size regardless of insulin treatment. The major findings from this diabetic cohort analysis are: (1) there were no major differences in the safety profile and efficacy between PES and SES in the overall cohort; (2) the “small-stent” group trended toward a higher incidence of MACE, and a higher rate of TLR and TVR rates compared with the “large stent” group. Nevertheless, we failed to find major differences between SES and PES in the incidence of the analyzed combined or single adverse events in the two stent size subgroups; (3) the incidence of TLR increased progressively along each year both with SES and PES, although the temporal course of TLR progression seems different between the two DES; (4) when subjects were stratified for insulin treatment, there were no significant differences between PES and SES in the incidence MACE, D/MI, TLR overall, regardless of the native stent size groups.
Our results are also comparable with some other larger registries. For example, Deamen et al. compared 2-year clinical outcomes in 708 patients with diabetes as part of the RESEARCH and T-SEARCH registries. They showed no significant differences between SES and PES with regard to D/MI, TVR [15]. In addition, the KOMATE registries, comparing 3-year outcomes in 634 diabetic patients, found no important difference in major adverse cardiac events between PES/SES [16]. Chiu et al. in 835 matched pair of diabetic patients treated by SES or PES as part of the Cardiac Care Network of Ontario registry found similar D/MI and TVR rates between the two types of stents at 3-year follow-up, that were comparable with our results [14]. Finally, Wolf et al. using the NHLBI Dynamic Registry, showed no significant differences between the two-stent types in any 2-year outcome of diabetic patients treated by PES (328) or SES (677) [17]. In our study, we confirmed and extended these results at a longer follow-up. Different from our results, a recent meta-analysis of five randomized clinical trials comparing SES with PES in diabetic patients demonstrated that SES were associated with a decreased risk for TLR compared with PES (hazard ratio 0.3, P < 0.001). However, it should be noticed that in this pooled analysis less than 1,200 patients were enrolled and clinical follow-up did not exceed the 9 months in three of the studies and the 24 months in the other two [18]. Similarly, two randomized controlled studies have reported that SES decreased MACE compared with PES in diabetic patients, but again these two studies were limited to only 8- and 9-month of clinical and angiographic follow-up, respectively [7, 19]. Furthermore, the patients enrolled in these randomized trials had lower risk profile when compared with our study population that includes older patients with higher rate of previous revascularization and of multivessel disease. On the contrary, our results are consistent to a previous meta-analysis of 35 randomized controlled trials by Stettler et al. who failed to found superiority of SES over PES, in term of restenosis and TLR rate in patients with diabetes throughout 4-year of follow-up [2].
Less reassuring, by our data, is the incidence of TLR that increased progressively along each year with both SES and PES, raising concerns on the delayed restenosis or late “catch up phenomenon” as demonstrated first in preclinical studies and then observed also in randomized clinical trials like SIRIUS and TAXUS IV at 5-year follow-up [20]. Of interest is also the slope of the TLR curve that appears to be milder in the SES group within 24 months and sharper thereafter, when compared with the PES group, mainly in the “large stent” subgroup. Although it might be due to chance, we cannot exclude a different time-course of restenosis as a function of the drug eluted by the stent, as shown also by others [21, 22]. This observation further highlights the importance of long-term follow-up data, particularly in complex subset much as diabetic patients, to make robust inferences on the safety and effectiveness of the new generations of DESs versus the first DESs generation or surgery. Of interest are also the potential biological and pharmacological explanations for the apparent differences in efficacy of the two DES types among diabetic and non-diabetic patients. Sirolimus- and paclitaxel-eluting stents provide local delivery of drugs that inhibits vascular smooth muscle cell proliferation, reducing therefore neointimal hyperplasia and restenosis. Sirolimus is a cytostatic agent that has the ability to inhibit the mammalian target of rapamycin regulatory protein kinase and blocks the cell cycle at the G1-S checkpoint. Mammalian target of rapamycin is regulated by growth factors via the phosphoinositide 3-kinase insulin signal transduction pathway, which is degraded in people with diabetes [23]. Paclitaxel, on the other hand, stabilizes microtubules and arrests the cell cycle at the G1 and G2/M junctions. The mechanism by which paclitaxel prevents restenosis is independent of the phosphoinositide 3-kinase pathway and is not affected by insulin resistance [23]. This would suggest that sirolimus might be less effective in diabetic patients. To note, Kereiakes et al. have recently shown that also everolimus elution provides comparable clinical benefit of paclitaxel among diabetic patients with reference vessel diameters ≥ 2.5 mm enrolled in the SPIRIT IV and XIENCE V [24].
The vessel size remains an important determinant of outcomes in DES era with the sirolimus appearing to be more effective than PES in reducing clinical measures of restenosis, particularly in small vessels (less able to accommodate neo-intimal hyperplasia) [9] and in non-diabetic patients [16]. Our data support the notion that the “small-stent” group have a higher rate of TLR and TVR rates compared with the “large stent” group, as shown also in non-diabetic patients by others [9, 10, 11]. Nevertheless, SES and PES were equally safe and efficacious in our population even after stratification for “small” versus “large” stent size and the treatment with insulin or not. Similarly to our results, a recent analysis of an unselected cohort of 1,253 diabetic patients from the REWARDS registry failed to found at 3-year follow-up differences between patients treated by insulin or not [25].
Limitation
We recognize that this study suffers from the inherent limitations of any observational study.
To account for the imbalances in characteristics of the treatment groups, we verified our multivariate Cox proportional hazards regression analysis on the large unmatched cohort by applying a propensity-score matched analysis. However, we cannot exclude there may still be some residual confounding by unmeasured factors. On the other hand, the fact that two independent statistical methods produces analogous results provide credence to our study findings. In addition, this study provides important information on a highly representative sample of “real world” patients who underwent PCI in routine clinical practice. Another limitation of this study is that we lacked QCA information on vessel size, in-stent lumen loss, in-stent thrombosis, and maximal balloon pressure-atm. However, clinically important safety end points such as myocardial infarction and death were analyzed and in a large population and we found no major clinically significant difference in the combined end-point of D/MI between the two stent types, supporting our statement of similar safety profiles between the two stents. There is no information on the serologic efficacy of diabetic treatment, including laboratory assessment of glycosylated hemoglobin or ophthalmologic evaluation for detection of microvascular abnormalities. The severity of diabetes was estimated only by analysis of insulin requirement. Finally, we did not formally assess the cost effectiveness of the two different stent types and associated outcomes.
CONCLUSIONS
In conclusion, in this large multicenter observational study of sirolimus and paclitaxel stents, we found that PES and SES are equally efficacious and have a similar safety profiles in diabetic patients undergoing PCI in clinical practice, regardless of the stent size and treatment with or without insulin throughout 1 to 5-year follow-up. Nevertheless, less reassuring is the observed increase in the incidence of target lesion revascularization along each year both in-patient receiving SES and PES, raising concerns on the delayed restenosis or late “catch up” as a clinical significant phenomenon. These results are significant in light of the emergence of new generations of DES. They should foster confidence in future comparison of the second generations of DES without concerns for the differential safety and efficacy between the two first generation DES in the diabetic population. On the contrary, they also highlight the importance of long-term follow-up results to understand whether or not the new generations of DES are really a step forward in diabetic patients when compared with the first generation DES or surgery.
REFERENCES
- 1, , , , , , , , , , , , , . Impact of drug-eluting stents among insulin-treated diabetic patients: A report from the National Heart, Lung, and Blood Institute Dynamic Registry. JACC Cardiovasc Interv 2008; 1: 139–147.
- 2, , , , , , , , , , , , , , , , , , , , , , , , , , , , . Outcomes associated with drug-eluting and bare-metal stents: A collaborative network meta-analysis. Lancet 2007; 370: 937–948.
- 3, , , , , , , , the REALITY Trial Investigators. Sirolimus- vs. paclitaxel-eluting stents in de novo coronary artery lesions: The REALITY trial: A randomized controlled trial. JAMA 2006; 295: 895–904.
- 4, , , , , , , , , , , , , , , , , . Sirolimus-eluting and paclitaxel-eluting stents for coronary revascularization. N Engl J Med 2005; 353: 653–662.
- 5, , , , . Robustness of late lumen loss in discriminating drug-eluting stents across variable observational and randomized trials. Circulation 2005; 112: 2833–2839.
- 6, , , , , , , , , ; the ISAR-DIABETES Study Investigators. Paclitaxel-eluting or sirolimus-eluting stents to prevent restenosis in diabetic patients. N Engl J Med 2005; 353: 663–670.
- 7, , , , , , , , , , , , , , , , , , , , . A randomized comparison of sirolimus- versus paclitaxel-eluting stent implantation in patients with diabetes mellitus. J Am Coll Cardiol 2008; 52: 727–733.
- 8, , , , , , , , , , , . Head-to-head comparison of sirolimus- and paclitaxel-eluting stent in the same diabetic patient with multiple coronary artery lesions: A prospective, randomized, multicenter study. Diabetes Care 2008; 31: 15–19.
- 9, , , , , , , , , , , , , , , , , , . Impact of vessel size on outcome after implantation of sirolimus-eluting and paclitaxel-eluting stents a subgroup analysis of the SIRTAX trial. J Am Coll Cardiol 2007; 50: 1123–1131.
- 10, , , , , , . Vessel size and outcome after coronary drug-eluting stent placement results from a large cohort of patients treated with sirolimus- or paclitaxel-eluting stents. J Am Coll Cardiol 2006; 48: 1304–1309.
- 11, , , , , , , , ; the BASKET Investigators. Long-term benefit–risk balance of drug-eluting vs. bare-metal stents in daily practice: Does stent diameter matter? Three-year follow-up of BASKET. Eur Heart J 2009; 30: 16–24.
- 12, , . A comparison of the ability of different propensity score models to balance measured variables between treated and untreated subjects: A Monte Carlo study. Stat Med 2007; 26: 734–753.
- 13, . Survival Analysis: Techniques for Censored and Truncated Data. New York: Springer-Verlag; 1997.
- 14, , , , , , , . Paclitaxel versus sirolimus stents in diabetic and nondiabetic patients. Circ Cardiovasc Qual Outcomes 2009; 2: 96–107.
- 15, , , , , , , . The long-term value of sirolimus- and paclitaxel-eluting stents over bare metal stents in patients with diabetes mellitus. Eur Heart J 2007; 28: 26–32.
- 16, , , , , , , , , , , , , , , , , , , , ; for the KOrean Multicenter Angioplasty TEam (KOMATE) Investigators. Comparison of sirolimus-eluting stent and paclitaxel- eluting stent for long-term cardiac adverse events in diabetic patients: The Korean multicenter angioplasty team (KOMATE) registry. Catheter Cardiovasc Interv 2008; 72: 601–607.
- 17, , , , , , , , , , , , , . Paclitaxel-eluting versus sirolimus-eluting stents in diabetes mellitus a report from the national heart, lung, and blood institute dynamic registry. Circ Cardiovasc Interv 2010; 3: 42–49.
- 18, , . Meta-analysis of five randomized clinical trials comparing sirolimus- versus paclitaxel-eluting stents in patients with diabetes mellitus. Am J Cardiol 2010; 105: 64–68.
- 19, , , , , , , , , . Comparison of the sirolimus-eluting versus paclitaxel-eluting coronary stent in patients with diabetes mellitus: The diabetes and drug-eluting stent (DiabeDES) randomized angiography trial. Am J Cardiol 2009; 103: 345–349.
- 20, , , . Temporal course of neointimal formation after drug-eluting stent placement: Is our understanding of restenosis changing? JACC Cardiovasc Interv 2009; 2: 300–302.
- 21, , , , , , , , , , , , , , , , , , , , . A randomized comparison of sirolimus- versus paclitaxel-eluting stent implantation in patients with diabetes mellitus. 4-Years clinical outcomes of DES-DIABETES (drug-eluting stent in patients with DIABETES mellitus) Trial. JACC Cardiovasc Interv 2011; 4: 310–316.
- 22, , , , ; ARTS I and II Investigators. 5-Years follow-up of coronary revascularization in diabetic patients with multivessel coronary artery disease. Insights from ARTS (arterial revascularization therapy study)-II and ARTS-I Trials. JACC Cardiovasc Interv 2011; 4: 317–323.
- 23. TOR of the cell cycle: Are there important implications for diabetics in the era of the drug-eluting stent? Catheter Cardiovasc Interv 2004; 61: 233–236.
- 24, , , , , , , , , , , , . Outcomes in diabetic and nondiabetic patients treated with everolimus- or paclitaxel-eluting stents: Results from the SPIRIT IV clinical trial (clinical evaluation of the XIENCE V everolimus eluting coronary stent system). J Am Coll Cardiol 2010; 56: 2084–2089.
- 25, , , , , , , , , , , , , , . Three-year outcomes following sirolimus- versus paclitaxel-eluting stent implantation in an unselected population with coronary artery disease (from the REWARDS Registry). Am J Cardiol 2010; 106: 504–510.
Supporting Information
Additional Supporting Information may be found in the online version of this article.
| Filename | Format | Size | Description |
|---|---|---|---|
| CCD_24445_sm_SuppInfo.pdf | 38K | Supporting Appendix. |
Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.