Author + information
- Received September 20, 2013
- Revision received November 29, 2013
- Accepted December 5, 2013
- Published online May 1, 2014.
- Kyung Woo Park, MD, PhD∗,
- Joo Myung Lee, MD, MPH∗,
- Si-Hyuck Kang, MD∗,
- Hyo-Suk Ahn, MD∗,
- Hyun-Jae Kang, MD, PhD∗,
- Bon-Kwon Koo, MD, PhD∗,
- Jay Young Rhew, MD, PhD†,
- Sun Ho Hwang, MD, PhD‡,
- Sung Yoon Lee, MD, PhD§,
- Tae Soo Kang, MD, PhD‖,
- Choong Hwan Kwak, MD, PhD¶,
- Bum-Kee Hong, MD, PhD#,
- Cheol Woong Yu, MD, PhD∗∗,
- In-Whan Seong, MD, PhD††,
- Taehoon Ahn, MD, PhD‡‡,
- Han Cheol Lee, MD, PhD§§,
- Sang Wook Lim, MD, PhD‖‖ and
- Hyo-Soo Kim, MD, PhD∗∗ ()
- ∗Department of Internal Medicine and Cardiovascular Center, Seoul National University Hospital, Seoul, Republic of Korea
- †Department of Internal Medicine and Cardiovascular Center, Presbyterian Medical Center, Jeonju, Republic of Korea
- ‡Department of Internal Medicine and Cardiovascular Center, Gwangju Veterans Hospital, Gwangju, Republic of Korea
- §Department of Internal Medicine and Cardiovascular Center, Inje University Ilsan Paik Hospital, Ilsan, Republic of Korea
- ‖Department of Internal Medicine and Cardiovascular Center, Dankook University Hospital, Cheonan, Republic of Korea
- ¶Department of Internal Medicine and Cardiovascular Center, Gyeongsang National University Hospital, Jinju, Republic of Korea
- #Department of Internal Medicine and Cardiovascular Center, Gangnam Severance Hospital, Seoul, Republic of Korea
- ∗∗Department of Internal Medicine and Cardiovascular Center, Sejong General Hospital, Sejong Heart Institute, Bucheon, Republic of Korea
- ††Department of Internal Medicine and Cardiovascular Center, Chungnam University Hospital, Daejeon, Republic of Korea
- ‡‡Department of Internal Medicine and Cardiovascular Center, Gachon University Gil Hospital, Incheon, Republic of Korea
- §§Department of Internal Medicine and Cardiovascular Center, Busan National University Hospital, Busan, Republic of Korea
- ‖‖Department of Internal Medicine and Cardiovascular Center, CHA Bundang Medical Center and CHA University, Seongnam, Republic of Korea
- ↵∗Reprint requests and correspondence:
Dr. Hyo-Soo Kim, Department of Internal Medicine, Seoul National University Hospital, 28 Yeongeon-dong, Chongno-gu, Seoul 110-744, Korea.
Objectives This study sought to compare everolimus-eluting stents (EES) versus Resolute zotarolimus-eluting stents (ZES) in terms of patient- or stent-related clinical outcomes in an “all-comer” group of patients with diabetes mellitus (DM) who underwent percutaneous coronary intervention.
Background DM significantly increases the risk of adverse events after percutaneous coronary intervention. The efficacy and safety of second-generation drug-eluting stents, in particular EES versus ZES, in patients with DM have not been extensively evaluated.
Methods Patients with DM (1,855 of 5,054 patients, 36.7%) from 2 prospective registries (the EXCELLENT [Efficacy of Xience/Promus Versus Cypher in Reducing Late Loss After Stenting] registry and RESOLUTE-Korea [Registry to Evaluate the Efficacy of Zotarolimus-Eluting Stent]) who were treated with EES (n = 1,149) or ZES (n = 706) were compared. Stent-related outcome was target lesion failure (TLF), and patient-oriented composite events were a composite of all-cause mortality, any myocardial infarction, and any revascularization.
Results Despite a higher risk patient profile in the ZES group, both TLF (43 of 1,149 [3.7%] vs. 25 of 706 [3.5%], p = 0.899) and patient-oriented composite events (104 of 1,149 [9.1%] vs. 72 of 706 [10.2%], p = 0.416) were similar between the EES and ZES in patients with DM at 1 year. In those without DM, EES and ZES also showed comparable incidence of TLF (39 of 1,882 [2.1%] vs. 33 of 1,292 [2.6%], p = 0.370) and patient-oriented composite events (119 of 1,882 [6.3%] vs. 81 of 1,292 [6.3%], p = 0.951), which were all significantly lower than in the DM patients. These results were corroborated by similar findings from the propensity score-matched cohort. Upon multivariate analysis, chronic renal failure was the most powerful predictor of TLF in DM patients (hazard ratio: 4.39, 95% confidence interval: 1.91 to 10.09, p < 0.001).
Conclusions After unrestricted use of second-generation drug-eluting stents in all-comers receiving percutaneous coronary intervention, both EES and ZES showed comparable clinical outcomes in the patients with DM up to 1 year of follow-up. DM compared with non-DM patients showed significantly worse patient- and stent-related outcomes. Nonetheless, overall incidences of TLF were low, even in the patients with DM, suggesting excellent safety and efficacy of both types of second-generation drug-eluting stents in this high-risk subgroup of patients.
Implantation of drug-eluting stents (DES) has demonstrated superiority in reducing the need for repeat revascularization compared with bare-metal stents in substudies of randomized controlled trials (RCT), a multicenter observational registry, and meta-analysis in patients with diabetes mellitus (DM) (1–4). Nonetheless, DM is still among the most important risk factors for adverse clinical events, even in the DES era (5,6). Second-generation DES have almost completely replaced first-generation DES, in the treatment of coronary artery disease. The representative second-generation DES, everolimus-eluting stents (EES) have shown noninferior or comparable clinical outcomes with first-generation DES in patients with DM (7,8). Another second-generation DES, the Resolute zotarolimus-eluting stent (ZES), recently received U.S. Food and Drug Administration labeling for use in patients with DM (9). However, head-to-head comparisons between EES and ZES have been limited to 2 representative randomized controlled trials (RESOLUTE All Comers [Randomized Comparison of a Zotarolimus-Eluting Stent With an Everolimus-Eluting Stent] trial and TWENTE [A Prospective Randomized Trial of Zotarolimus-Eluting Stents and Everolimus-Eluting Stents in Patients With Coronary Artery Disease]) (10,11). Recently, we published the 1-year comparison of EES versus ZES in 5,054 patients from 2 multicenter registries (EXCELLENT [Efficacy of Xience/Promus Versus Cypher in Reducing Late Loss After Stenting] registry and RESOLUTE-Korea [Registry to Evaluate the Efficacy of Zotarolimus-Eluting Stent]) (12). To the best of our knowledge, there are limited data regarding direct comparison of EES with ZES in patients with DM. Therefore, we compared EES versus ZES in terms of patient- and stent-related clinical outcomes in all-comers with DM undergoing percutaneous coronary intervention (PCI) with second-generation DES.
An extended description of study methods are presented in the Online Appendix.
Study design and patient population
This study evaluated 1-year clinical outcomes of EES (Xience V, Abbott Vascular, Abbott Park, Illinois) and zotarolimus-eluting stents (ZES) (Resolute, Medtronic Cardiovascular, Minneapolis, Minnesota) in the patients with DM from the EXCELLENT and RESOLUTE-Korea registries, which enrolled all-comers treated with ≥1 EES or ZES (3,056 patients in 29 participating centers and 1,998 patients in 25 participating centers, respectively) without exclusions (12). Among the 5,054 patients, 36.7% (n = 1,855) had type 2 DM and were treated with EES (1,149 of 3,056 patients, 37.9%) or ZES (706 of 1,998 patients, 35.3%). The flow of patients in the study is presented in Figure 1.
After index PCI, follow-ups were performed at 1, 3, 9, and 12 months; angiography was optional at 9 months. All relevant medical records were reviewed for any clinical event and adjudicated by an external clinical event committee. Using the Korean health system's unique identification numbers, the vital status of all patients was cross-checked. The study protocol was approved by the ethics committee at each participating center and was conducted according to the principals of the Declaration of Helsinki. All patients provided written informed consent.
Definitions and outcome analysis
DM (type 1 or type 2) was defined as either a previous diagnosis of DM treated with pharmacologic or nonpharmacologic measure, or a new DM was defined according to the American Diabetes Association as history of either presence of classic symptoms of DM with unequivocal elevation of plasma glucose (2 h post-prandial or random of >200 mg/dl) or fasting plasma glucose elevation on >2 occasions of ≥126 mg/dl during hospitalization. Patients were considered insulin-treated if they were taking insulin. Patients were considered non-insulin-treated if they were taking only oral hypoglycemic agents or were on a therapeutic lifestyle modification only or both oral agents and therapeutic life-style modification. The primary stent-related clinical outcome was target lesion failure (TLF), defined as cardiac death, target vessel myocardial infarction (MI), or clinically indicated target lesion revascularization (TLR) by percutaneous or surgical methods at 12 months. All clinical outcomes were defined according to the Academic Research Consortium (13,14). The key secondary outcomes were patient-oriented composite events, which included all-cause mortality, any MI, and any revascularization. Other secondary outcomes included individual components of the primary and key secondary clinical outcomes; target or non–target vessel MI; clinically driven or angiographically driven repeat revascularization including TLR or target vessel revascularization (13); and stent thrombosis (ST) defined according to the Academic Research Consortium as definite, probable, or possible (13,14). The indication of PCI was considered “off-label” if any of the following features were present: serum creatinine concentration ≥140 μmol/l (1.6 mg/dl); left ventricular ejection fraction <30%; an acute MI within the previous 72 h; >1 lesion per vessel; ≥2 vessels treated with a stent; a lesion length ≥28 mm; or a bifurcated lesion, bypass graft, in-stent restenosis, unprotected left main coronary artery, presence of thrombus, or total occlusion (15,16). Chronic renal failure was defined as an estimated glomerular filtration rate <60 ml/min calculated by means of the Cockcroft-Gault formula.
The analysis was performed in 2 parts. First, analysis and comparison of primary and secondary clinical outcomes were conducted in the crude population. Second, a propensity score matched population was selected to adjust for uneven distribution of baseline characteristics. Comparison of treatment effect between 2 stent groups was stratified by DM versus non-DM. Categorical variables were presented as numbers and relative frequencies (percentages) and were compared using the chi-square test or the Fisher exact test for independent groups and a 2-tail p value. Normally distributed continuous variables were expressed as means ± SD and were analyzed using the independent sample t test. Kaplan-Meier analysis was used to calculate cumulative incidence of primary and secondary clinical outcomes, and log-rank test or Breslow test was used to compare between-group differences, as appropriate. Because differences in baseline clinical and angiographic characteristics could affect the primary and secondary clinical outcomes, a 1:1 matched analysis without replacement was performed using propensity scores. The propensity scores were estimated with multiple logistic regression analysis that included all covariates listed in Tables 1 and 2, except outcome variables. Logistic regression model was separately conducted to generate a propensity score, which was the probability that a patient received a ZES or which was the probability that a patient had DM. The discrimination and calibration abilities of the propensity score model were assessed by C-statistics (0.652 for the first model according to the stent types and 0.705 for the second model according to the presence or not of DM) and the Hosmer-Lemeshow statistics (chi-square: 8.375, p = 0.398 and chi-square: 7.107, p = 0.525, respectively). The matching processes were also performed separately to compare the clinical outcomes between EES and ZES, or between patients with or without DM. For matching, nearest neighbor matching with a caliper width of 0.6 SD was used because this value has been shown to eliminate over 90% of the bias in the observed confounders (17). Success of the propensity score matching was assessed by calculating the percentage of standardized differences of the baseline characteristics. A <10% difference supports the assumption of a balance between matched groups (18). A Cox proportional hazard regression model was used to identify independent predictors of primary clinical outcome TLF. The covariates used in multivariate analysis were selected if they were significantly different between the 2 groups (p < 0.1) or if they had predictive values. All probability values were 2-sided and p values < 0.05 were considered statistically significant. The statistical package SPSS (version 18.0, SPSS Inc., Chicago, Illinois) and R programming language (version 2.15.2, R Foundation for Statistical Computing, Vienna, Austria) were used for statistical analyses.
Baseline patient and angiographic characteristics of diabetes population
Among the overall population of both registries (n = 5,054), the final analysis population was 1,855 patients with DM with 2,688 lesions (EES: 1,149 patients with 1,660 lesions, ZES: 706 patients with 1,028 lesions), and 3,174 nondiabetic patients with 4,365 lesions (EES: 1,882 patients with 2,557 lesions, ZES: 1,292 patients with 1,808 lesions), after exclusion of the patients with unknown status of DM. In the overall population of 5,054 patients, 55 (1.1%) in the EES group and 32 (0.6%) patients in the ZES group, respectively, were lost to follow-up before the 12-month follow-up visit; however, all were confirmed alive. Baseline clinical and angiographic characteristics are presented in Tables 1 and 2, respectively. Among the 1,855 patients, 209 patients (11.3%) were newly diagnosed with DM according to the criteria of the American Diabetes Association at the time of index procedure. The proportion of newly diagnosed DM was not different between the 2 stent groups (12.2% in the EES group vs. 9.8% in the ZES group, p = 0.113). Most of the patients with DM (72.8%) were on oral hypoglycemic agents. Of the 1,855 patients with diabetes, 255 patients (13.7%) required insulin treatment. Glycemic control, quantified by hemoglobin A1C, was significantly poorer in insulin-treated than in non-insulin-treated patients (8.1 ± 1.8% vs. 7.4 ± 1.3% for insulin- vs. non-insulin-treated diabetics, p < 0.001). Regarding baseline risk factors, both stent groups were largely comparable. However, the proportion of insulin-treated patients was higher (12.2% vs. 16.3%, p = 0.015) and glycemic control was worse (hemoglobin A1C: 7.4 ± 1.5% vs. 7.6 ± 1.4%, p = 0.031) in the ZES group. Although the proportion of dyslipidemia was higher in the ZES group, the mean low-density cholesterol level was similar between the 2 groups. In addition, the ZES group had a slightly higher baseline risk, including higher frequency of ST-segment elevation MI patients, bifurcation lesions, and longer stent length. PCI was also more often off-label in the ZES group. However, type B2 or C lesions were more common in the EES group (Tables 1 and 2). Overall, acute coronary syndrome accounted for 60.3%, and PCI was off-label in 77.7% of the cases. Patients with multivessel disease accounted for 62.8%, and the mean stent length was 41.1 ± 26.7 mm per patient, all of which were significantly higher in the patients with diabetes than in nondiabetic patients in both stent groups. The large proportion of high-risk patients and lesions suggests that our study patients were an enriched PCI population, reflecting real-world practice in Korea without any exclusion or restriction. In addition, 37.1% of the cases were performed under intravascular ultrasound guidance. Despite the large proportion of high-risk patients and lesions, the device, lesion, and procedure success rates were excellent for both stents and did not show between-group differences (Table 2).
Clinical outcomes at 1 year in crude population
In diabetic patients, the TLF rate was 3.7% (43 of 1,149 patients) in the EES group and 3.5% (25 of 706 patients) in the ZES group (relative risk [RR]: 0.95, 95% confidence interval [CI]: 0.58 to 1.54, p = 0.899). The rates of the individual components of TLF, cardiac death, TLR, and target vessel–related MI were also similar between the 2 stent groups (Table 3). The rate of patient-oriented composite events was similar as well (9.1% vs. 10.2%, RR: 1.13, 95% CI: 0.85 to 1.50, p = 0.416), as were the individual components (all-cause death, any revascularization, any MI). Approximately one-half of the cases of target vessel MI (5 of 11, 45.5%) were due to ST (Table 3). In nondiabetic patients, the incidence of TLF was also comparable between the 2 stent groups (39 of 1,882 [2.1%] vs. 33 of 1,292 [2.6%], p = 0.370), as with patient-oriented composite events (119 of 1,882 [6.3%] vs. 81 of 1,292 [6.3%], p = 0.951). The rates of TLF and patient-oriented composite events were both significantly higher in diabetics versus nondiabetics, regardless of stent type (Fig. 2). However, both stent groups did not show any differences in the cumulative incidence of TLF or patient-oriented composite events, both in the diabetic and nondiabetic patients (Fig. 2). The survival curves of individual components of TLF or patient-oriented composite events are presented in Online Figure 1.
The incidence of Academic Research Consortium–defined definite or probable ST through 1 year was very low, regardless of diabetes: the rates of ST were 0.5% (10 of 1,855 patients) for the diabetic patients and 0.5% (15 of 3,174 patients) in the nondiabetic patients. In diabetic patients, there was a numerically higher rate of definite or probable ST in the EES group without statistical significance (0.8% vs. 0.1%, p = 0.100) (Table 4). However, the incidence of ST did not show any clustering of events in nondiabetic patients (0.5% vs. 0.5%, p = 0.956). All diabetic patients with ST were on dual antiplatelet therapy, except in 1 case in which both agents were discontinued due to spontaneous subdural hemorrhage 2 days after PCI. A detailed description of all ST cases in the patients with DM is presented in Online Table 1.
Notably, the subgroup of insulin-treated patients (n = 255) showed similar rates of TLF and patient-oriented composite events compared with the non-insulin-treated patients (TLF: 4.3% vs. 3.6%, p = 0.589, patient-oriented composite events: 10.2% vs. 9.4%, p = 0.730 for insulin- vs. non-insulin-treated patients, respectively). Conversely, patients with chronic renal failure showed significantly higher incidence of both TLF (9.3% vs. 3.3%, p < 0.01) and patient-oriented composite events (14.7% vs. 9.2%, p = 0.038) (Fig. 3).
Propensity score matched group analysis
Propensity score matching, according to the type of stents, yielded 934 patients (467 pairs) with more balanced baseline characteristics (Online Table 2). All standardized differences of the 44 adjusted variables were <10% (Online Fig. 2). At 1 year, TLF occurred in 3.9% and 3.4% (RR: 0.89, 95% CI: 0.46 to 1.72, p = 0.862) and patient-oriented composite events in 9.4% and 8.6% (RR: 0.91, 95% CI: 0.60 to 1.37, p = 0.732) in the EES and ZES groups, respectively. The individual components of both TLF and patient-oriented composite events, and the incidence of definite and probable ST were also similar (Online Table 3). In the survival analysis, cumulative incidences of TLF or patient-oriented composite events were comparable between the EES and ZES groups, regardless of the presence of diabetes (Fig. 2).
To minimize the risk of bias in comparing diabetic and nondiabetic patients, both groups were matched according to the presence of DM. As a result, 2,774 patients were matched (1,387 patients with or without DM, respectively) with more balanced baseline characteristics (Online Table 4). As expected, except for significantly higher hemoglobin A1C level in the diabetes group, standardized differences of baseline characteristics were <10% (Online Fig. 2). Both TLF (3.6% in diabetic patients vs. 1.9% in nondiabetic patients, respectively, RR: 1.85, 95% CI: 1.17 to 2.94, p = 0.008) and patient-oriented composite events (9.0% vs. 6.2%, RR: 1.45, 95% CI: 1.12 to 1.89, p = 0.005) were significantly higher in the patients with diabetes than in the nondiabetic patients (Online Table 5). In addition, the patients with DM showed significantly higher incidence of target vessel MI and target vessel revascularization (Online Table 5).
Independent predictors of target lesion failure in the patients with diabetes mellitus
In the multivariate Cox regression model, chronic renal failure was the strongest predictor of TLF (adjusted hazard ratio [HR]: 4.393, 95% CI: 1.913 to 10.09, p < 0.001). Other significant predictors of TLF included in-stent restenosis, left main coronary artery PCI, and small vessel intervention, but not the type of stent itself (HR: 0.92, 95% CI: 0.48 to 1.79, p = 0.810) (Table 5, Online Table 4). The overall Harrell C-index of the model was 0.780 (95% CI: 0.699 to 0.861).
This is the first comprehensive registry analysis evaluating the safety and efficacy of EES (Xience V EES) versus ZES (Resolute ZES) in the patients with DM. Our results demonstrated that EES and ZES showed comparable results regarding stent- and patient-related composite outcomes at 1 year in the patients with DM, both in the crude and propensity score matched populations. Although the incidences of TLF or patient-oriented composite events were significantly higher in the patients with DM than in the nondiabetic patients, the overall incidence was low, suggesting excellent safety and efficacy of both types of second-generation DES in this high-risk subgroup of patients. In addition, the patient-related outcomes were about 3-fold higher than the stent-related outcomes (TLF) were, suggesting the importance of secondary prevention and integrated medical management of comorbidities along with DM, including diabetic nephropathy, hypertension, and peripheral arterial disease. Finally, the chronic renal failure was the strongest predictor of TLF within 1 year in the patients with DM.
The safety and efficacy of the second-generation DES in the patients with DM have been continuously evaluated. The EES was shown to be noninferior to sirolimus-eluting stent for in-segment late loss at 8 months in the ESSENCE-DIABETES (Randomized Comparison of Everolimus-Eluting Stent Versus Sirolimus-Eluting Stent Implantation for De Novo Coronary Artery Disease in Patients With Diabetes Mellitus) trial (7). Although underpowered for clinical outcomes, death, MI, and TLR were not significantly different between EES (Xience V) and sirolimus-eluting stents. Also post-hoc subgroup analysis from 4 pooled RCT, comparing EES versus paclitaxel-eluting stent, showed no difference in clinical outcomes after 2 years of follow-up between EES and paclitaxel-eluting stents (8). On the other hand, ZES (Resolute) have showed significantly lower incidence of target vessel failure at 1 year (7.8%) than the pre-defined DES performance goal of 14.5% in the patients with DM and achieved the first U.S. Food and Drug Administration approval for PCI in patients with diabetes (9). Although both of the second-generation DES, compared with first generation DES, have shown improved clinical outcomes in the diabetic patients, the data regarding head-to-head comparison of these stents in the patients with DM have been limited. The representative direct comparisons of EES versus ZES (i.e., Resolute All-Comers and TWENTE trials) did not focus on the diabetic subgroup. The proportion of the patients with DM was limited to 538 patients (23.4%) in the Resolute All-Comers trial and 301 patients (21.6%) in the TWENTE trial, which were relatively underpowered to compare clinical outcomes in the diabetic population. In addition, it is well known that patients with higher-risk profiles and higher early in-hospital mortality tend to be excluded from participation in clinical trials, even with all-comers design (19). We evaluated 1,855 diabetics with 2,688 lesions, and the majority of patients had ≥1 off-label indication (77.7%), and there was no restriction or exclusion criteria regarding disease severity or lesion complexity. In addition, over 98% of the patients were strictly followed, and the survival status of all patients was thoroughly investigated. In this regard, this analysis of prospective observational registries has the strength of including a broader patient population with quite a large sample size, which is more reflective of everyday practice, compared with former RCT.
Interestingly, insulin-treated patients showed numerically higher but statistically insignificant rates of TLF and patient-oriented composite events. Previous trials using sirolimus-eluting stents (SIRIUS [Sirolimus-Coated Bx Velocity Balloon-Expandable Stent in the Treatment of Patients With De Novo Coronary Artery Lesions] and DIABETES [Diabetes and Sirolimus-Eluting Stent]) have shown conflicting results regarding whether insulin-treated patients show increased risk of adverse outcome (1,2). Also, pooled analysis of EES versus PES from the SPIRIT (Clinical Evaluation of the Xience V Everolimus Eluting Coronary Stent System in the Treatment of Patients With De Novo Native Coronary Artery Lesions) II, III, IV and the COMPARE (Cozaar in Marfan Patients Reduces Aortic Enlargement) trials showed no significant difference in 2-year hard clinical endpoints (cardiac death, MI, or ST) according to insulin treatment; however, significant interaction existed between diabetes treatment and stent type (8). Conversely, recent pooled analysis of patient-level data from the RESOLUTE Global Clinical Program showed significantly higher incidence of target vessel failure (16.4% versus 10.4%, p = 0.02) or TLF (13.5% versus 7.9%, p = 0.01), mainly driven by higher incidence of cardiac death or target vessel MI (3.9% vs. 8.6%, p = 0.01), in the insulin-treated patients after 2 years following ZES implantation (9). Whether insulin-dependency significantly affects outcome of PCI with second-generation DES needs to be clarified in larger-scale data.
On multivariate analysis, several factors were found to significantly increase the risk of TLF including chronic renal failure, in-stent restenosis, left main vessel PCI, and small vessel diameter. Chronic renal failure and left main vessel PCI have been well-recognized risk factors for major adverse cardiac events (20) and angiographic restenosis (21) after stent implantation in both the bare-metal stent era and the first-generation DES era. Small vessel stenting may lead to increased risk of periprocedural myocardial necrosis and has been reported to be a risk factor for ST (22). Because diabetic patients tend to have smaller vessel diameters, this could adversely affect the long-term clinical outcomes, as was corroborated in our results. Treatment of in-stent restenosis requires stent implantation in an area that has previously been stented. This can result in injury zone mismatch, a gap between the stents, fracture of the stent, polymer disruption, or a combination of these mechanisms (23). Thus, treatment of in-stent restenosis lesions still remains a challenge even in the DES era. All things considered, however, the low rates of both stent- and patient-oriented outcomes in the present analysis reaffirms that the development of a new generation of DES has been in the right direction. Nonetheless, more investigation about long-term clinical outcomes after PCI with second-generation DES in insulin-treated diabetic patients is needed.
First, this was a nonrandomized comparison of 2 different registries. Therefore, this study cannot be free from inherent limitations of observational registries such as allocation bias and uneven distribution of risk factors. Although we used propensity score matching to minimize the allocation bias and control for potential confounding variables, the possibilities of uncontrolled and unknown confounding factors need to be considered. Second, analysis of clinical outcome was limited to 1 year after index PCI. Our study is not able to make any conclusions regarding long-term prognosis over 1 year in diabetic patients. Further follow-up is required, especially to address safety issues such as ST. Third, because data were from observational registries, the clinical events may not have been captured with scrutiny, and patient follow-up may not have been as strict as would be in a randomized trial. This may have been the reason for the low event rates, especially the rate of target lesion–related MI, which was much lower in our study than in the previous RCT and pooled analysis of EES and ZES (8–11). We believe that this may be due to the fact that systematic collection of cardiac enzymes after PCI is not the routine practice in Korea. Some centers even discharge their patients on the same day after PCI via the radial approach. Therefore, we believe the true rate of MI would have been much higher if the cardiac enzymes were followed systematically. Although we cross-checked the vital status of 100% of the study patients with the Korean national database using a citizen registration number that is unique to each individual, we cannot exclude the possibility of under-reporting of clinical outcomes other than death such as MI, TLR, or most importantly, ST, in the patients who were lost to follow-up but are still alive. Lastly, although we thoroughly collected baseline and discharge medication data including dual antiplatelet agents, we did not include cilostazol as 1 of the parameters to collect, therefore, the clinical efficacy of triple antiplatelet agent in diabetic patients could not be evaluated.
After unrestricted use of second-generation DES in all-comers receiving PCI, both EES and ZES showed comparable clinical outcomes in the patients with DM up to 1 year of follow-up. The patients with DM showed significantly worse patient- and stent-related outcomes than did the nondiabetic patients. Nonetheless, overall incidences of target lesion failure were low, even in the patients with DM, suggesting excellent safety and efficacy of both types of second-generation DES in this high-risk subgroup of patients.
For supplemental study methods, figures, and tables, please see the online version of this paper.
This study was supported by a grant from the Innovative Research Institute for Cell Therapy, Seoul National University Hospital (#A062260), sponsored by the Ministry of Health and Welfare, Republic of Korea. The 2 registries were also funded from unrestricted grants from Abbott Vascular Republic of Korea and Medtronic Republic of Korea. The funding sources of the study had no role in study design, data collection, monitoring, analysis, interpretation, or writing of the paper. Dr. H.-S. Kim has received unrestricted grants and lecture honoraria from Abbott Vascular Korea and Medtronic Republic of Korea. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Park and J. M. Lee contributed equally to this study.
- Abbreviations and Acronyms
- confidence interval
- drug-eluting stent(s)
- diabetes mellitus
- everolimus-eluting stent(s)
- hazard ratio
- myocardial infarction
- percutaneous coronary intervention
- randomized controlled trial(s)
- relative risk
- stent thrombosis
- target lesion failure
- target lesion revascularization
- zotarolimus-eluting stent(s)
- Received September 20, 2013.
- Revision received November 29, 2013.
- Accepted December 5, 2013.
- American College of Cardiology Foundation
- Moussa I.,
- Leon M.B.,
- Baim D.S.,
- et al.
- Sabaté M.,
- Jiménez-Quevedo P.,
- Angiolillo D.J.,
- et al.,
- for the DIABETES Investigators
- Ortolani P.,
- Balducelli M.,
- Marzaroli P.,
- et al.
- Hong S.J.,
- Kim M.H.,
- Ahn T.H.,
- et al.
- Kastrati A.,
- Massberg S.,
- Ndrepepa G.
- Kim W.J.,
- Lee S.W.,
- Park S.W.,
- et al.,
- for the ESSENCE-DIABETES Study Investigators
- Stone G.W.,
- Kedhi E.,
- Kereiakes D.J.,
- et al.
- Silber S.,
- Serruys P.W.,
- Leon M.B.,
- et al.
- Silber S.,
- Windecker S.,
- Vranckx P.,
- et al.,
- for the RESOLUTE All Comers Investigators
- Tandjung K.,
- Sen H.,
- Lam M.K.,
- et al.
- Park K.W.,
- Lee J.M.,
- Kang S.H.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- von Birgelen C.,
- Basalus M.W.,
- Tandjung K.,
- et al.
- Gu X.S.,
- Rosenbaum P.R.
- Cohen J.
- de Boer S.P.,
- Lenzen M.J.,
- Oemrawsingh R.M.,
- et al.
- Saltzman A.J.,
- Stone G.W.,
- Claessen B.E.,
- et al.
- Brodie B.,
- Pokharel Y.,
- Garg A.,
- et al.