We compared the effectiveness of drug-eluting stents (DESs) to bare-metal stents (BMSs) in ostial lesions from an unrestricted patient cohort with 3-year follow-up. DESs have proved more effective at decreasing repeat revascularization rates compared to BMSs in patients with uncomplicated coronary artery disease. Whether DESs provide similar benefits in ostial lesions is not clearly defined. We analyzed data from 775 patients in the National, Heart, Lung, and Blood Institute Dynamic Registry undergoing stenting of ostial lesions with DESs or BMSs. Patients were followed for 3 years for the occurrence of myocardial infarction (MI), repeat revascularization (coronary bypass surgery/repeat percutaneous coronary intervention), and death. In total 439 patients had 464 ostial lesions treated with BMSs and 336 patients had 351 ostial lesions treated with DESs. Adjusted DES versus BMS 3-year hazard ratios were 1.03 (95% confidence interval 0.60 to 1.78, p = 0.90) for death, 1.40 (0.83 to 2.37, p = 0.21) for MI, and 0.81 (0.59 to 1.11, p = 0.19) for repeat revascularization. In patients undergoing percutaneous coronary intervention for aorto-ostial disease (n = 200), death and repeat revascularization did not differ between stent types, but DES-treated patients had more MI during follow-up. For coronary ostial disease (n = 574), 3-year observed rates of death or MI did not differ; however, repeat revascularization was more common in the BMS group. In conclusion, use of DESs for ostial lesions was associated with no difference in the hazard of death, MI, or overall rates of repeat revascularization compared to BMS use.
Drug-eluting stents (DESs) have proved more effective than bare-metal stents (BMSs) in decreasing the need for repeat revascularization. Complex lesions, however, have generally been excluded from initial randomized comparisons. As a result, the effectiveness of DESs compared with BMSs in complex coronary lesions including ostial lesions is less clear. Ostial lesions present a unique challenge given the higher prevalence of calcification, turbulent blood flow patterns, rigidity, elastic recoil, and ability to achieve correct stent placement compared to nonostial lesions. Furthermore, aorto-ostial lesions, representing aortic wall disease, are a unique subset of ostial lesions where the pathology of ostial lesion is different. Previous studies comparing DESs to BMSs in ostial lesions are limited in the number of patients studied, location of lesions, and duration of follow-up. The purpose of this report is to describe 3-year outcomes after unrestricted use of DESs versus BMSs in ostial coronary lesions from the National, Heart, Lung, and Blood Institute (NHLBI) Dynamic Registry.
Methods
This dynamic registry is a multicenter NHLBI-sponsored prospective observational study of consecutive patients undergoing percutaneous coronary intervention (PCI) at selected centers in North America. It is composed of 5 “waves” of patient enrollment, each enrolling >2,000 patients since 1997, with the intent to study changes in PCI technology over time. Waves 1 to 3 enrolled patients when only BMSs were available. Waves 4 (2004) and 5 (2006) enrolled patients during the DES era. To decrease election bias, BMS-treated patients were selected only from waves 1 to 3.
Trained research coordinators collected demographic, clinical, angiographic, and procedural data pertaining to the index PCI procedure and vital status, repeat hospitalization, and medication use information during follow-up using standardized report forms. Hospital charts and coronary angiograms were reviewed to assess inpatient outcomes. Follow-up data were collected at 1 month, 6 months, and annually thereafter by direct patient contact. Patients enrolled in waves 1 and 3 were followed for 1 year and follow-up for patients in waves 2, 4, and 5 was extended. Routine follow-up angiography was not performed and staged PCI was not considered repeat PCI. Lesion-specific data were collected to determine target vessel revascularization rates.
Death was included as all-cause mortality. Other end points evaluated were myocardial infarction (MI) and any repeat revascularization (PCI or any coronary artery bypass grafting after index PCI). MI was defined as the presence of ≥2 of the following findings: typical chest pain lasting 20 minutes and not relieved by nitroglycerin; serial electrocardiograms showing changes from baseline in ST and T waves and/or Q waves in >2 contiguous leads; increase in creatine kinase to >2 times upper limit of normal with a creatine kinase-MB index of >5%; and increase in troponin to >2 times upper limit of normal. Ostial lesions were defined as aorto-ostial lesions (right coronary artery, left main coronary artery, saphenous vein graft, or arterial graft ostial lesions) or coronary ostial lesions within the coronary tree of >50% stenosis severity by visual assessment.
Continuous variables were compared by Student’s t or Wilcoxon nonparametric tests and categorical variables by chi-square or Fisher’s exact tests. Three-year cumulative event rates were estimated with the Kaplan–Meier method, and unadjusted survival curves were compared using log-rank statistic. Cox proportional hazards model was used to estimate 3-year hazard ratios (HRs) for clinical events in relation to stent type. Covariate adjustment was performed with demographic, clinical, and lesion/procedural characteristics entered into outcome-specific models including an indicator variable for stent type. For each outcome potential confounders were adjusted for in a forward stepwise manner to determine the final model, and variables with a p value <0.10 were included in age-adjusted final models. Proportionality assumptions for Cox models were met. Patients not developing an event of interest by 3 years for those enrolled in waves 2, 4, and 5 were censored at 3 years and the same was done at 1 year for patients enrolled in waves 1 and 3. A 2-sided p value <0.05 was considered significant for all statistical analyses. We performed a subgroup analysis evaluating clinical event rates in patients treated with aorto-ostial and coronary ostial lesions.
Results
Baseline characteristics comparing differences between BMS- and DES-treated patients are presented in Table 1 . The 2 groups were of similar age and had a similar proportion of woman patients. DES-treated patients were more likely to have a history of diabetes, hypertension, hypercholesterolemia, and previous PCI. BMS-treated patients were more likely to present with acute coronary syndromes and cardiogenic shock and less likely to receive periprocedural thienopyridine therapy.
| Variable | BMS (n = 439) | DES (n = 336) | p Value |
|---|---|---|---|
| Mean age (years) | 65.5 | 65.4 | 0.86 |
| Women | 39.2% | 35.7% | 0.32 |
| Body mass index (kg/m 2 ) | 28.3 | 29.7 | <0.001 |
| Diabetes mellitus | 30.9% | 39.7% | 0.01 |
| Insulin therapy | 10.4% | 16.4% | 0.01 |
| Hypertension | 69.6% | 84.0% | <0.001 |
| Hypercholesterolemia | 65.7% | 83.6% | <0.001 |
| Smoker | 0.79 | ||
| Current | 18.8% | 18.8% | |
| Former | 43.3% | 45.5% | |
| Previous myocardial infarction | 37.5% | 27.5% | 0.004 |
| Previous angioplasty | 29.1% | 40.2% | <0.001 |
| Previous coronary bypass | 28.3% | 32.0% | 0.54 |
| Co-morbidities | |||
| Cerebrovascular | 8.3% | 11.4% | 0.15 |
| Renal insufficiency | 7.2% | 9.3% | 0.28 |
| Peripheral arterial disease | 12.9% | 11.1% | 0.43 |
| Ejection fraction, mean (%) | 51.1 | 51.9 | 0.27 |
| Number of coronary arteries diseased | 0.06 | ||
| 1 | 32.3% | 23.5% | |
| 2 | 30.3% | 34.5% | |
| 3 | 37.1% | 41.7% | |
| Mean left main coronary artery stenosis >50% | 12.1% | 22.0% | <0.001 |
| Reason for revascularization | |||
| Myocardial infarction | 20.1% | 18.5% | 0.57 |
| Unstable angina | 50.0% | 35.4% | <0.001 |
| Stable angina | 21.0% | 27.4% | 0.04 |
| Cardiogenic shock | 3.9% | 0.6% | 0.003 |
| Periprocedural medications | |||
| Thienopyridines | 61.3% | 86.6% | <0.001 |
| Heparin | 96.6% | 65.8% | <0.001 |
| Low-molecular-weight heparin (waves 2–5) | 2.9% | 3.3% | 0.79 |
| Glycoprotein IIb/IIIa inhibitor | 41.5% | 36.3% | 0.15 |
| Discharge medications | |||
| Aspirin | 91.1% | 98.5% | <0.001 |
| Angiotensin-converting enzyme inhibitor | 38.8% | 51.2% | <0.001 |
| β Blocker | 64.7% | 84.1% | <0.001 |
| Calcium channel blocker | 26.9% | 16.8% | <0.001 |
| Statins | 56.3% | 83.8% | <0.001 |
| Thienopyridines | 89.2% | 99.1% | <0.001 |
| Mean number of lesions | 3.7 | 3.9 | 0.20 |
| Mean lesions attempted | 1.8 | 1.5 | 0.001 |
| Procedural success | 96.1% | 98.2% | 0.09 |
| Mean stents/patient | 1.85 | 1.83 | 0.82 |
| Insurance status | 0.075 | ||
| Medicare | 50.3% | 42.4% | |
| Public | 11.8% | 12.2% | |
| Private | 36.0% | 41.5% | |
| Self | 1.8% | 3.9% |
In total 439 patients had 464 ostial lesions attempted with BMSs compared to 351 ostial lesions attempted with 336 DESs ( Table 2 ). At the lesion level, the location of lesions and reference vessel diameter were similar between the 2 groups. Lesions treated with DESs were significantly longer, treated with more stents/lesion, and more likely to be class C compared to BMS-treated lesions. BMS-treated lesions were more likely to be thrombotic, which is consistent with higher rates of acute coronary syndromes in BMS-treated patients. Overall angiographic and procedural success rates were high and similar for the DES and BMS groups.
| Variable | BMS (n = 464) | DES (n = 351) | p Value |
|---|---|---|---|
| Location | 0.43 | ||
| Left main coronary artery | 5.6% | 7.7% | |
| Left anterior descending coronary artery | 36.9% | 34.8% | |
| Left circumflex coronary artery | 21.2% | 25.1% | |
| Right coronary artery | 24.2% | 22.2% | |
| Bypass graft | 12.1% | 10.3% | |
| Mean reference vessel diameter (mm) | 3.2 | 3.2 | 0.72 |
| Mean lesion length (mm) | 11.7 | 15.5 | <0.001 |
| Complex lesion types | |||
| Total occlusion | 8.4% | 6.8% | 0.41 |
| Thrombus present | 12.8% | 7.8% | 0.02 |
| Calcified lesion | 36.4% | 40.5% | 0.24 |
| Ulcerated lesion | 10.5% | 11.4% | 0.70 |
| American College of Cardiology/American Heart Association classification | <0.001 | ||
| A | 4.6% | 1.7% | |
| B1 | 17.3% | 22.6% | |
| B2 | 52.7% | 37.4% | |
| C | 25.4% | 38.3% | |
| Sirolimus-eluting stent | 65.8% | ||
| Paclitaxel-eluting stent | 34.2% | ||
| Procedural complications | |||
| Major dissection | 5.2% | 2.3% | 0.03 |
| Perforation | 0.4% | 0.0% | 0.22 |
| Embolization | 1.3% | 1.4% | 0.87 |
| Side branch occlusion | 1.9% | 2.3% | 0.74 |
| Angiographic success | 97.8% | 98.0% | 0.87 |
| Stents/lesion (mean) | 1.17 | 1.28 | 0.006 |
Rate of in-hospital death (BMS 2.5% vs DES 0.6%, p = 0.04) was higher in the BMS group. Rates of MI (BMS 4.8% vs DES 4.5%, p = 0.83), coronary artery bypass grafting (BMS 1.1% vs DES 0.0%, p = 0.05), stroke (BMS 0.9% vs DES 0.3%, p = 0.29), and bleeding requiring transfusion (BMS 3.9% vs DES 2.1%, p = 0.15) were similar between groups. DES-treated patients were more likely to receive dual antiplatelet therapy, statins, β blockers, and angiotensin-converting enzyme inhibitors on discharge, possibly reflecting different practice patterns during BMS and DES recruitment phases.
At 3-year follow-up, observed rates of death, MI, and repeat revascularization did not differ ( Table 3 and Figure 1 ). In patients undergoing repeat procedures, repeat PCI did not differ; however, coronary artery bypass grafting was significantly higher in the BMS group. After adjustment, there was no difference in rate of death, MI, and repeat revascularization ( Figure 2 ). Overall 3-year target vessel revascularization rates from available data were not different for the DES versus BMS groups.
| Variable | BMS (n = 439) | DES (n = 336) | p Value |
|---|---|---|---|
| All-cause death | 15.30% | 13.20% | 0.12 |
| Myocardial infarction | 9.90% | 12.30% | 0.67 |
| Repeat percutaneous coronary intervention | 16.30% | 23.80% | 0.68 |
| Coronary artery bypass grafting | 10.40% | 5.30% | 0.002 |
| Coronary artery bypass grafting/repeat percutaneous coronary intervention (repeat revascularization) | 24.20% | 26.40% | 0.17 |
| Target vessel revascularization | 9.30% (137) ⁎ | 7.80% (314) | 0.6 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
