The outcomes for patients undergoing percutaneous coronary interventions (PCI) with drug-eluting stents (DESs) and bare metal stents (BMSs) have been compared in many studies for patients with ST-segment elevation myocardial infarction. However, little is known about the relative outcomes for patients with non-ST-segment elevation myocardial infarction (NSTEMI). The aim of the present study was to compare the NSTEMI outcomes for PCI with DESs and BMSs. New York’s PCI registry was used to propensity-match 4,776 pairs of patients with NSTEMI who had received DESs and BMSs from January 1, 2003 to December 31, 2007. These patients were followed up through December 31, 2008 to test for differences in mortality, target vessel revascularization, and total repeat revascularization. The outcomes were also compared for various patient subsets. At a median follow-up period of 3.68 years, the patients receiving DESs had significantly lower mortality (16.58% vs 14.52%, difference 2.06%, p <0.001), target vessel revascularization (13.08% vs 11.04%, p = 0.009), and total repeat revascularization (22.16% vs 18.77%, p <0.001). The patients receiving paclitaxel-eluting and sirolimus-eluting stents both experienced superior outcomes compared to patients receiving BMSs. The patients receiving DESs had significantly lower mortality rates than their propensity-matched counterparts receiving BMSs when they were ≥65 years (difference 2.29%, p = 0.01) and male (difference 2.77%, p = 0.003). In conclusion, patients with NSTEMI undergoing PCI experienced lower 4-year mortality, target vessel revascularization, and repeat revascularization rates when they had received DESs than when they had received BMSs, and patients who were >65 years old, and men received notable benefits.
Numerous studies have compared the outcomes for drug-eluting stents (DESs) and bare metal stents (BMSs) and for several high-risk subgroups, including acute myocardial infarction. However, nearly all the studies comparing DESs and BMSs for patients with acute myocardial infarction have been limited to the subset of patients with ST-segment elevation myocardial infarction and have not included patients with non–ST-segment elevation myocardial infarction (NSTEMI).
The purpose of the present study was to use a population-based clinical registry of patients presenting with NSTEMI to compare the outcomes according to whether the patients had received DESs or BMSs. The outcomes included all-cause mortality, target vessel percutaneous coronary intervention (PCI), and repeat revascularization. Because the study was an observational study, we propensity-matched the patients who had received DESs and BMSs using a large variety of variables, including patient demographics, nature and extent of coronary heart disease, admission and transfer status, interval since the onset of symptoms, hemodynamic status, congestive heart failure, several co-morbidities, and target vessel and number of vessels treated.
Methods
The data were obtained from the New York State’s Percutaneous Coronary Interventions Reporting System, a mandatory registry in New York, initially developed in 1991, that contains detailed information for each patient undergoing PCI in New York state on the demographics; preprocedural risk factors; periprocedural complications; types of devices used; extent of disease and lesions treated; admission, discharge, and procedure dates; discharge disposition and destination; and hospital and operator identifiers. These data are matched to the New York administrative data, the Statewide Planning and Research Cooperative System, and are audited by the New York State Department of Health’s utilization review agent to ensure completeness and accuracy.
For the present study, the data were matched to the Statewide Planning and Research Cooperative System to confirm that the myocardial infarction was NSTEMI using the International Classification of Disease, Ninth Revision, Clinical Modification code 410.71. The Percutaneous Coronary Interventions Reporting System data were also matched to New York’s vital statistics data; thus, these patients could be followed up after discharge for evidence of subsequent death. Also, data from New York’s Cardiac Surgery Reporting System was used with the Percutaneous Coronary Interventions Reporting System data to identify repeat target vessel revascularizations after discharge through December 31, 2008.
All patients who had experienced an NSTEMI within 7 days before undergoing stenting with either BMSs or DESs (but not both) from January 1, 2003 to December 31, 2007 were included in the present study, except for patients who had undergone thrombolysis treatment within 7 days before the index PCI (n = 354), who had undergone PCI >7 days after symptom onset (n = 1,568), had undergone coronary artery bypass grafting in the same admission after PCI (n = 45), or were from out of state (n = 801). All other patients with NSTEMI who had undergone stenting from January 1, 2003 to December 31, 2007 (5,062 BMS patients and 14,086 DES patients, total 19,148) were followed up through December 31, 2008. Of the 14,086 DES patients, 8,168 had received sirolimus-eluting stents and 6,071 paclitaxel-eluting stents.
The end points included all-cause mortality (median follow-up 3.67 years for all patients, 3.71 years for DES patients, and 3.62 years for BMS patients), target vessel PCI, and repeat revascularization at 6 months and for each of the 4 years after the index procedure. Mortality was censored in calculating the rates for the other 2 outcomes.
The differences between the patients with NSTEMI undergoing stenting with BMSs and DESs in the prevalence of various patient risk factors (e.g., demographics, co-morbidities, left ventricular function, hemodynamic state, diseased vessels, interval since the onset of symptoms) and the differences in-hospital/30-day mortality were tested using chi-square tests and Fisher’s exact tests.
Because the patients were not randomized to DES versus BMS, we identified the patient preprocedural characteristics that were potentially related to the type of stent received and matched-paired patients according to those characteristics using propensity-matched samples.
The matching characteristics included demographics, body surface area, admission type, interval since the onset of symptoms, anatomic group (number of diseased vessels and whether the patient had proximal or nonproximal left anterior descending disease, left ventricular function, hemodynamic status, congestive heart failure, several co-morbidities, target vessel attempted, number of vessels treated, and date range of the procedure.
These characteristics were used to develop a logistic regression model that predicted the probability that a given patient with NSTEMI would receive DESs. This value, the propensity score, was used to match patients without replacement on a 1-to-1 basis to minimize the overall distance in propensity scores between the groups. To ensure that the factors with the greatest risk (e.g., hemodynamic instability, low ejection fraction, renal failure, age ≥75 years) were not the cause of selection bias, each BMS patient with one or more of these factors was matched to a DES patient with the same characteristics before applying the propensity score. Also, the DES and BMS patients were matched exactly according to the period of stenting (January 1, 2003 to December 31, 2004, January 1, 2005 to June 30, 2006, and July 1, 2006 to December 31, 2007). Differences between the 2 matched samples in the prevalence of the propensity model variables were tested using standardized differences in the observed prevalence of the variables in the matched groups. The propensity-matched pairs were then used to analyze the differences in outcomes between the 2 groups. The methods were used to compare the DESs and BMSs for each of the outcomes, taking into account that the samples were matched.
The mortality rates for DESs and BMSs were compared at 1, 2, 3, and 4 years for 2 preselected patient characteristics (age ≥65 years and gender) and for the type of DES (paclitaxel-eluting stents vs BMSs and sirolimus-eluting stents vs BMSs). All tests were 2-sided and conducted at the 0.05 level, and all analyses were conducted using SAS, version 9.1 (SAS Institute).
Results
A total of 19,148 patients with NSTEMI were used to identify propensity-matched groups of DES and BMS patients, with 14,086 (73.6%) receiving DESs and 5,062 BMSs ( Table 1 ). In general, the DES patients were younger, were less likely to be white or black, had greater body surface areas, had different intervals from symptom onset to balloon time, were less likely to have left main disease, had different patterns of diseased vessels/proximal left anterior descending disease, had greater ejection fractions, had a lower prevalence of congestive heart failure and cerebrovascular and peripheral vascular disease, were more likely to be diabetic or receiving renal dialysis, were less likely to have had previous cardiac surgery, and were more likely to have had multivessel PCI at the index admission.
| Variable | All Patients (n = 19,148) | DES (n = 14,086) | BMS (n = 5,062) | p Value |
|---|---|---|---|---|
| In-hospital/30-day mortality | 1.48 | 1.31 (185) | 1.96 (99) | 0.001 |
| Age (years) | <0.001 | |||
| ≤54 | 25.08 | 25.95 | 22.66 | |
| 55–64 | 24.92 | 25.42 | 23.55 | |
| 65–74 | 23.29 | 23.07 | 23.90 | |
| ≥75 | 26.70 | 25.55 | 29.89 | |
| Women | 35.04 | 35.12 | 34.83 | 0.71 |
| Race | <0.001 | |||
| White | 83.67 | 83.29 | 84.73 | |
| Black | 10.45 | 10.43 | 10.51 | |
| Other | 5.88 | 6.28 | 4.76 | |
| Body surface area <1.7 m 2 | 12.02 | 11.64 | 13.10 | 0.01 |
| Emergency admission | 11.83 | 11.78 | 11.97 | 0.72 |
| Duration of acute coronary syndrome | <0.001 | |||
| 0–11 hours | 6.38 | 6.16 | 7.01 | |
| 11–23 hours | 11.28 | 11.19 | 11.52 | |
| 1–3 days | 61.93 | 63.01 | 58.95 | |
| 4–7 days | 20.41 | 19.65 | 22.52 | |
| Left main artery disease | 3.96 | 3.56 | 5.06 | |
| Coronary arteries narrowed | <0.001 | |||
| 2, no proximal left anterior descending | 24.92 | 25.00 | 24.71 | |
| 3, no proximal left anterior descending | 11.34 | 11.42 | 11.12 | |
| 1, proximal left anterior descending | 8.74 | 9.23 | 7.39 | |
| 2, proximal left anterior descending | 8.37 | 8.70 | 7.45 | |
| 3, proximal left anterior descending | 8.09 | 7.90 | 8.63 | |
| Ejection fraction (%) | <0.001 | |||
| ≤30 | 5.22 | 4.95 | 5.94 | |
| 30–39 | 9.29 | 8.99 | 10.11 | |
| 40–49 | 20.56 | 20.41 | 20.96 | |
| ≥50 | 58.78 | 59.88 | 55.75 | |
| Missing | 6.15 | 5.76 | 7.23 | |
| Hemodynamically, unstable | 0.44 | 0.38 | 0.59 | 0.05 |
| Shock | 0.09 | 0.06 | 0.18 | 0.02 |
| Congestive heart failure | 13.20 | 12.65 | 14.74 | <0.001 |
| Cerebrovascular disease | 8.26 | 7.85 | 9.38 | <0.001 |
| Peripheral vascular disease | 6.91 | 6.62 | 7.72 | 0.01 |
| Chronic obstructive pulmonary disease | 7.73 | 7.65 | 7.98 | 0.44 |
| Diabetes mellitus | 28.73 | 29.27 | 27.22 | 0.01 |
| Renal dysfunction | 0.002 | |||
| Creatinine >2.5 mg/dl, no dialysis | 1.66 | 1.48 | 2.15 | |
| Renal dialysis | 2.43 | 2.54 | 2.13 | |
| Previous percutaneous coronary intervention | 17.30 | 17.56 | 16.57 | 0.11 |
| Previous surgery | 14.09 | 12.84 | 17.54 | <0.001 |
| Target coronary artery | <0.001 | |||
| Left main artery | 0.46 | 0.43 | 0.57 | |
| Right coronary artery | 24.29 | 22.82 | 28.37 | |
| Left anterior descending artery | 32.74 | 33.80 | 29.79 | |
| Left circumflex artery | 26.45 | 25.51 | 29.06 | |
| >1 Vessel attempted | 16.05 | 17.44 | 12.21 |
The data from the 9,552 patients who were propensity-matched into 2 paired groups of 4,776 patients, 1 of whom received DESs and 1 of whom received BMSs are listed in Table 2 . Note that of the 5,062 BMS patients listed in Table 1 , all but 286 were propensity-matched with a DES patient. After matching, the percentage standardized difference between the propensity-matched groups was very low for most risk factors, with only 1 risk factor >5.0% and all well below the recommended maximum of 10.0%.
| Risk Factor | All Patients (n = 9,552) | DES (n = 4,776) | BMS (n = 4,776) | Standardized Difference |
|---|---|---|---|---|
| In-hospital/30-day mortality | 1.54 | 1.51 (72) | 1.57 (75) | 0.80 |
| Age (years) | ||||
| ≤54 | 24.16 | 24.66 | 23.66 | 2.35 |
| 55–64 | 23.65 | 23.14 | 24.16 | 2.41 |
| 65–74 | 24.03 | 24.04 | 24.02 | 0.05 |
| ≥75 | 28.16 | 28.16 | 28.16 | 0.00 |
| Women | 34.79 | 34.82 | 34.76 | 0.13 |
| Race | ||||
| White | 84.18 | 83.63 | 84.74 | 3.04 |
| Black | 10.89 | 11.35 | 10.43 | 2.96 |
| Other | 4.93 | 5.03 | 4.84 | 0.87 |
| Body surface area <1.7 m 2 | 12.33 | 12.23 | 12.44 | 0.64 |
| Emergency admission | 12.02 | 12.44 | 11.60 | 2.58 |
| Duration of acute coronary syndrome | ||||
| 0–11 hours | 6.74 | 6.55 | 6.93 | 1.50 |
| 11–23 hours | 11.74 | 12.02 | 11.45 | 1.76 |
| 1–3 days | 60.39 | 61.33 | 59.44 | 3.85 |
| 4–7 days | 21.14 | 20.10 | 22.17 | 5.08 |
| Left main artery disease | 4.27 | 4.40 | 4.15 | 1.24 |
| Coronary arteries narrowed | ||||
| 2, no proximal left anterior descending | 24.70 | 24.48 | 24.92 | 1.02 |
| 3, no proximal left anterior descending | 10.75 | 10.70 | 10.80 | 0.34 |
| 1, proximal left anterior descending | 7.54 | 7.43 | 7.64 | 0.79 |
| 2, proximal left anterior descending | 7.29 | 7.29 | 7.29 | 0.00 |
| 3, proximal left anterior descending | 7.81 | 7.85 | 7.77 | 0.31 |
| Ejection fraction (%) | ||||
| ≤29 | 4.69 | 4.69 | 4.69 | 0.00 |
| 30–39 | 9.53 | 9.53 | 9.53 | 0.00 |
| 40–49 | 20.95 | 20.23 | 21.67 | 3.55 |
| ≥50 | 58.62 | 59.34 | 57.89 | 2.93 |
| Missing | 6.22 | 6.22 | 6.22 | 0.00 |
| Hemodynamically, unstable | 0.08 | 0.08 | 0.08 | 0.00 |
| Shock | 0.00 | 0.00 | 0.00 | 0.00 |
| Congestive heart failure | 12.98 | 12.65 | 13.32 | 1.99 |
| Cerebrovascular disease | 8.44 | 8.21 | 8.67 | 1.66 |
| Peripheral vascular disease | 7.10 | 7.24 | 6.95 | 1.14 |
| Chronic obstructive pulmonary disease | 6.64 | 6.64 | 6.64 | 0.00 |
| Diabetes mellitus | 26.86 | 27.22 | 26.51 | 1.61 |
| Renal dysfunction | ||||
| Creatinine >2.5 mg/dl, no dialysis | 1.07 | 1.07 | 1.07 | 0.00 |
| Renal dialysis | 1.53 | 1.53 | 1.53 | 0.00 |
| Previous percutaneous coronary intervention | 16.45 | 16.48 | 16.42 | 0.17 |
| Previous surgery | 15.66 | 15.56 | 15.77 | 0.58 |
| Target coronary artery | ||||
| Left main artery | 1.50 | 1.55 | 1.44 | 0.86 |
| Right coronary artery | 34.95 | 34.63 | 35.26 | 1.32 |
| Left anterior descending artery | 38.74 | 39.13 | 38.34 | 1.63 |
| Left circumflex artery | 37.23 | 37.02 | 37.44 | 0.87 |
| >1 Vessel attempted | 10.85 | 10.62 | 11.08 | 1.48 |
Table 3 and Figures 1 to 3 present the outcomes for all the outcome measures at 6 months and every year after the procedure through 4 years. The mortality rates favored DESs, with a difference of 1.32% after 6 months (4.25% vs 2.93%, p <0.001) that remained significant in favor of DESs in every subsequent period.
| Outcome | BMS (n = 4,776) | DES (n = 4,776) | Difference | p Value ⁎ |
|---|---|---|---|---|
| Interval to death (mo) | ||||
| 6 | 4.25% (203) | 2.93% (140) | 1.32% | <0.001 |
| 12 | 6.32% (302) | 4.73% (226) | 1.59% | <0.001 |
| 24 | 9.57% (436) | 7.51% (346) | 2.06% | <0.001 |
| 36 | 13.22% (551) | 11.20% (467) | 2.02% | <0.001 |
| 48 | 16.58% (647) | 14.52% (561) | 2.06% | <0.001 |
| Interval to first target vessel revascularization (mo) | ||||
| 6 | 5.37% (250) | 3.11% (146) | 2.26% | <0.001 |
| 12 | 7.29% (337) | 4.87% (227) | 2.42% | <0.001 |
| 24 | 9.46% (418) | 7.07% (317) | 2.39% | <0.001 |
| 36 | 11.17% (465) | 10.05% (405) | 1.12% | 0.038 |
| 48 | 13.08% (511) | 11.04% (429) | 2.04% | 0.009 |
| Interval to total revascularization † (mo) | ||||
| 6 | 9.23% (429) | 5.91% (277) | 3.32% | <0.001 |
| 12 | 12.26% (567) | 9.15% (426) | 3.11% | <0.001 |
| 24 | 16.11% (711) | 12.92% (580) | 3.19% | <0.001 |
| 36 | 19.14% (793) | 17.41% (711) | 1.73% | 0.004 |
| 48 | 22.16% (864) | 18.77% (743) | 3.39% | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
