Drug-eluting stents effectively inhibit neointimal hyperplasia within the first year, thereby reducing the need for repeat revascularization. However, a delayed pattern of restenosis might be more prominent in drug-eluting stents compared to bare metal stents (BMSs). The extent of restenosis of paclitaxel-eluting stents (PESs) long term after implantation in acute ST-segment elevation myocardial infarction is currently unknown. The present study was designed to evaluate very late luminal loss (VLLL) of PESs used in ST-segment elevation myocardial infarction compared to BMSs. A total of 116 patients (61 with PESs and 55 with BMSs) initially included in the Paclitaxel Eluting Stent Versus Conventional Stent in ST-segment Elevation Myocardial Infarction (PASSION) trial and who were free from previous lesion failure underwent angiographic follow-up. Off-line quantitative coronary analysis of the angiogram immediately after stent implantation and at follow-up was performed. The primary end point was VLLL within the stent. The presence of binary restenosis was defined as diameter stenosis >50% as a secondary end point. The mean interval between stent implantation and follow-up was 4.1 ± 0.5 years in both stent groups. In-stent VLLL was 0.12 mm (interquartile range −0.03 to 0.42) in the PES group versus 0.30 mm (interquartile range 0.08 to 0.69) in the BMS group (p = 0.011). In-segment binary restenosis was found in 4 patients (6.6%) with a PES and 6 patients (10.9%) with a BMS (p = 0.40). In conclusion, angiographic follow-up 4 years after implantation in ST-segment elevation myocardial infarction showed that in patients prospectively randomized to PESs or BMSs, VLLL was low in both stent groups. PESs were associated with lower VLLL than BMSs, and the observed rate of binary restenosis was not significantly different between the 2 stent groups.
The extent of restenosis very late after implantation of drug-eluting stents in primary percutaneous coronary intervention (PPCI) is currently not known. To elucidate the very late effects of paclitaxel-eluting stents (PESs) compared to bare metal stents (BMSs) in patients presenting with acute ST-segment elevation myocardial infarction (STEMI), we performed an angiographic follow-up study 4 years after implantation in patients prospectively randomized to either a PES or a BMS with the same stent platform.
Methods
In 2003 and 2004, a total of 619 patients were included in the Paclitaxel-Eluting Stent Versus Conventional Stent in Acute ST-Elevation Myocardial Infarction (PASSION) trial, which was conducted in 2 high-volume centers in The Netherlands (Onze Lieve Vrouwe Gasthuis, Amsterdam, and St. Antonius Hospital, Nieuwegein). The patients presenting with STEMI were prospectively randomized to receive either a PES or a BMS during PPCI. The details of the study design and the results of the clinical follow-up have both been published previously.
For the present follow-up study, we interviewed all patients who were free of target lesion failure and asked them if they were willing to undergo repeat angiography. The exclusion criteria were renal impairment at risk of contrast nephropathy defined as creatinine >130 mmol/L (or >1.47 mg/dl), stenting in the saphenous vein graft at inclusion, subsequent target vessel revascularization by percutaneous coronary intervention or coronary artery bypass grafting, the inability to provide written informed consent, angiogram not suitable for quantitative coronary angiography, and participation in another clinical trial. The institutional ethics committees of both participating centers approved the present study.
At the Onze Lieve Vrouwe Gasthuis, coronary angiography was performed routinely using the transradial approach. At the St. Antonius Hospital, the transfemoral approach was the standard procedure. After intracoronary injection of 100 to 200 μg nitroglycerin, angiography of identical orthogonal views that showed the stent area free of foreshortening or vessel overlap, using a 6F guiding catheter, was performed. Quantitative coronary angiography of each procedure was performed off-line and independently by 2 investigators (M.V. and F.v.N.) who were unaware of patient data and stent type, using the QAngio XA, version 7.1, computerized edge-detection system (Medis, Leiden, The Netherlands). The angiographic end points were calculated as the average values of the analyses of both observers. In the case of disagreement between the 2 investigators defined as a difference of >2 standard deviations, a third investigator (G.A.) was consulted to achieve conformity. Measurements were made within the stent and within the stented segment, which included 5 mm of its proximal and distal borders. Very late luminal loss (VLLL) was defined as the change in the minimal lumen diameter determined by laboratory measurement from immediately after stent implantation to follow-up. Binary restenosis was present if the diameter stenosis was >50%. Qualitative assessment of the pattern of restenosis was performed according to the classification introduced by Mehran et al. In-stent restenosis (ISR) was divided into focal (≤10 mm in length), diffuse (>10 mm within the stent), proliferative (>10 mm extending outside the stent), and occlusive ISR.
The primary end point was in-stent VLLL. The secondary end points included in-segment VLLL, the presence of binary restenosis, minimal luminal diameter, change in the percentage of diameter stenosis, and percentage of diameter stenosis at follow-up.
The data are presented as the rates and percentages for the categorical variables and as the mean value ± SD for normally distributed, continuous variables or the median with the interquartile range (IQR), when applicable. The differences between groups were tested with the Student t test or Mann-Whitney U test for continuous variables and the chi-square test or Fisher’s exact test for categorical data. The differences in the distribution of continuous variables between both groups were compared using the Kolmogorov-Smirnov test. For all tests, the differences were considered significant if the 2-sided p value was <0.05. Data processing and statistical analysis was performed using the Statistical Package Social Sciences, version 18 (SPSS, Chicago, Illinois) statistical program for personal computers.
The study size was determined by the size of the original study and the fraction of patients who were willing to undergo repeat angiography. With the inclusion of 2 times 60 patients, we calculated that with a 2-sided test for differences in average VLLL with an α level of 0.05, the study would have 90% power to detect a 0.30-mm reduction in PESs compared to BMS. In the absence of comparable studies, we estimated this required sample size according to the Treatment of De Novo Coronary Disease Using a Single Paclitaxel-Eluting Stent (TAXUS) IV trial, which found a reduction of late loss of 0.53 mm in PESs versus BMSs at 9 months and a SD of 0.50 in PES.
Results
Of a total of 619 patients included in the PASSION trial, 516 met the clinical inclusion criteria as determined at 2 years of follow-up. Of the 516 patients, 117 gave written informed consent. Of these patients, 8 had angiograms that were clinically indicated because of nontarget vessel coronary disease. The most common reason for not entering the study was refusal to undergo coronary angiography in the absence of complaints. One patient was excluded from additional analysis, because the initial angiogram was not suitable for quantitative coronary angiography.
The baseline clinical, angiographic, and procedural characteristics of the 61 patients who had received a PES and the 55 patients who had received a BMS are listed in Table 1 . The baseline characteristics were well matched. The mean age was 56 ± 10 years in the PES group and 55 ± 9 years in the BMS group (p = 0.55). At follow-up, the use of medication and the level of secondary prevention were similar in both groups ( Table 2 ). Of all the patients, excluding those who underwent clinically indicated angiography because of nontarget vessel disease, 103 (95.4%) reported Canadian Cardiovascular Society class I angina; 5 patients (4.6%) experienced Canadian Cardiovascular Society class II angina.
| Variable | PES (n = 61) | BMS (n = 55) | p Value |
|---|---|---|---|
| Age (years) | 56 ± 10 | 55 ± 9 | 0.55 |
| Men | 49 (80%) | 47 (86%) | 0.47 |
| Hypertension | 20 (33%) | 14 (26%) | 0.39 |
| Current smoker | 35 (57%) | 34 (62%) | 0.63 |
| Diabetes mellitus | 4 (7%) | 3 (6%) | 0.80 |
| Hypercholesterolemia ⁎ | 15 (25%) | 20 (36%) | 0.17 |
| Infarct-related artery | 0.17 | ||
| Left anterior descending | 37 (61%) | 26 (47%) | |
| Right | 20 (33%) | 20 (36%) | |
| Left circumflex | 4 (7%) | 9 (16%) | |
| Thrombolysis In Myocardial Infarction flow grade before procedure | 0.67 | ||
| 0 | 39 (64%) | 34 (63%) | |
| 1 | 5 (8%) | 3 (6%) | |
| 2 | 5 (8%) | 8 (15%) | |
| 3 | 12 (20%) | 9 (17%) | |
| Multivessel disease | 23 (38%) | 24 (44%) | 0.52 |
| Number of stents | 1.24 ± 0.50 | 1.19 ± 0.52 | 0.59 |
| Size stent (mm) | 3.25 ± 0.29 | 3.26 ± 0.37 | 0.82 |
| Stent length (mm) | 18.4 ± 5.4 | 19.6 ± 5.2 | 0.21 |
| Thrombus aspiration | 33 (54%) | 27 (49%) | 0.59 |
| Glycoprotein IIb/IIIa inhibitor | 51 (84%) | 41 (75%) | 0.23 |
| Maximum creatine kinase-MB (IU/L) | 184 ± 167 | 181 ± 156 | 0.94 |
| Procedural success | 59 (97%) | 55 (100%) | 0.18 |
⁎ Diagnosis previously made by physician or receiving lipid-lowering therapy.
| Variable | PES (n = 61) | BMS (n = 55) | p Value |
|---|---|---|---|
| Medication | |||
| Aspirin | 57 (93%) | 50 (91%) | 0.61 |
| Clopidogrel | 0 | 3 (6%) | 0.07 |
| Oral anticoagulation | 4 (7%) | 2 (4%) | 0.48 |
| β Blocker | 45 (74%) | 40 (73%) | 0.90 |
| Angiotensin-converting enzyme inhibitor/angiotensin receptor blocker | 37 (61%) | 29 (53%) | 0.39 |
| Statin | 56 (92%) | 49 (89%) | 0.62 |
| Calcium channel blocker | 6 (10%) | 7 (13%) | 0.62 |
| Nitrate | 8 (13%) | 4 (7%) | 0.30 |
| Diuretic | 10 (16%) | 8 (15%) | 0.78 |
| Canadian Cardiovascular Society angina class | |||
| I | 54 (98%) | 49 (92%) | 0.20 |
| II | 1 (2%) | 4 (8%) | |
| Stopped smoking | 10/35 (29%) | 7/34 (21%) | 0.44 |
| Cholesterol | 0.13 | ||
| mmol/L | 4.2 ± 0.9 | 4.5 ± 1.1 | |
| mg/dl | 162 ± 35 | 174 ± 43 | |
| High-density lipoprotein | 0.31 | ||
| mmol/L | 1.4 ± 0.5 | 1.3 ± 0.4 | |
| mg/dl | 54 ± 19 | 50 ± 15 | |
| Low-density lipoprotein | 0.24 | ||
| mmol/L | 2.2 ± 0.6 | 2.4 ± 1.0 | |
| mg/dl | 85 ± 23 | 93 ± 39 |
The initial procedural success, defined as Thrombolysis In Myocardial Infarction (TIMI) flow grade 3 and <50% residual stenosis, was achieved in 97% and 100% of patients in the PES and BMS groups, respectively (p = 0.18). At 3.18 ± 0.31 mm in the PES and 3.22 ± 0.36 mm in BMS groups (p = 0.49), the reference diameter after stent implantation was similar in both groups ( Table 3 ). The minimal lumen diameter immediately after intervention was comparable in both groups (2.71 ± 0.38 and 2.77 ± 0.39 in PES and BMS, respectively; p = 0.37).
| Variable | After Intervention | Follow-up | ||||
|---|---|---|---|---|---|---|
| PES (n = 61) | BMS (n = 55) | p Value | PES (n = 61) | BMS (n = 55) | p Value | |
| Interval after primary percutaneous coronary intervention (days) | 1,507 ± 172 | 1,519 ± 180 | 0.69 | |||
| Thrombolysis In Myocardial Infarction flow grade 3 | 59 (96.7%) | 55 (100%) | 0.18 | 61 (100%) | 55 (100%) | 1.00 |
| Reference diameter (mm) | 3.18 ± 0.31 | 3.22 ± 0.36 | 0.61 | 3.19 ± 0.31 | 3.20 ± 0.38 | 0.88 |
| Minimal luminal diameter (mm) | ||||||
| In-stent | 2.71 ± 0.38 | 2.77 ± 0.39 | 0.37 | 2.46 ± 0.58 | 2.37 ± 0.61 | 0.37 |
| In-segment | 2.32 ± 0.46 | 2.37 ± 0.40 | 0.41 | 2.28 ± 0.55 | 2.23 ± 0.58 | 0.63 |
| Degree of stenosis (%) | ||||||
| In-stent | 14.5 (10.0–20.2) | 13.1 (9.7–19.2) | 0.47 | 20.6 (12.8–28.5) | 24.5 (17.5–34.2) | 0.08 |
| In-segment | 26.4 (17.9–33.4) | 25.3 (21.0–33.4) | 0.70 | 28.2 (18.6–35.2) | 27.3 (22.2–37.4) | 0.47 |
| Late loss (mm) | ||||||
| In-stent | 0.12 (0.03–0.42) | 0.30 (0.08–0.69) | 0.011 | |||
| In-segment | 0.05 ± 0.46 | 0.15 ± 0.46 | 0.25 | |||
| Binary restenosis | ||||||
| In-stent | 3 (4.9%) | 4 (7.3%) | 0.60 | |||
| In-segment | 4 (6.6%) | 6 (10.9%) | 0.40 | |||
| Pattern of restenosis | 0.14 | |||||
| Focal | 3 (4.9%) | 2 (3.6%) | ||||
| Diffuse | 0 | 4 (7.3%) | ||||
| Proliferative | 1 (1.6%) | 0 | ||||
The mean interval from stent implantation to repeat angiography was 4.1 ± 0.5 years in the PES group and 4.1 ± 0.5 years in the BMS group (p = 0.69). The angiographic data at follow-up are listed in Table 3 . The average in-stent VLLL was significantly lower in the PES group compared to the BMS group at 0.12 mm (IQR −0.03 to 0.42) versus 0.30 mm (IQR 0.08 to 0.69), respectively (p = 0.011; Figure 1 ) . In-stent VLLL was not normally distributed, particularly in the PES group. The difference in the distribution of VLLL between the PES and BMS groups was statistically significant (p = 0.012, Kolmogorov-Smirnov test; Figures 2 and 3 ) . Accordingly, the distribution of the change in the percentage of diameter stenosis between implantation and follow-up was right-skewed. With an average of 20.6% (IQR 12.8 to 28.5) in-stent stenosis in the PES group and 24.5% (IQR 17.5 to 34.2) in-stent stenosis in the BMS group (p = 0.08), no significant difference was found in the percentage of diameter stenosis at follow-up ( Figure 4 ) . In contrast to in-stent VLLL, the mean in-segment VLLL was comparable in the PES and BMS group at 0.05 ± 0.46 mm and 0.15 ± 0.46 mm, respectively (p = 0.25). The mean in-stent minimal lumen diameter at follow-up was equivalent at 2.46 ± 0.58 mm in the PES group and 2.37 ± 0.61 mm in the BMS group (p = 0.37). In-segment binary restenosis was observed in 4 patients (6.6%) with the PES and in 6 patients (10.9%) with the BMS (p = 0.40). The pattern of restenosis was focal in 3 patients in the PES group (75% of patients with binary restenosis), and diffuse restenosis was seen in 4 patients (66.7%) in the BMS group (p = 0.14).
