The most common causes of in-stent restenosis (ISR) are intimal hyperplasia and stent under expansion. The purpose of this study was to use intravascular ultrasound (IVUS) to compare the ISR mechanisms of bare metal stents (BMS), first-generation drug-eluting stents (DES), and second-generation DES. There were 298 ISR lesions including 52 BMS, 73 sirolimus-eluting stents, 52 paclitaxel-eluting stents, 16 zotarolimus-eluting stents, and 105 everolimus-eluting stent. Mean patient age was 66.6 ± 1.1 years, 74.2% were men, and 48.3% had diabetes mellitus. BMS restenosis presented later (70.0 ± 66.7 months) with more intimal hyperplasia compared with DES (BMS 58.6 ± 15.5%, first-generation DES 52.6 ± 20.9%, second-generation DES 48.2 ± 22.2%, p = 0.02). Although reference lumen areas were similar in BMS and first- and second-generation DES, restenotic DES were longer (BMS 21.8 ± 13.5 mm, first-generation DES 29.4 ± 16.1 mm, second-generation DES 32.1 ± 18.7 mm, p = 0.003), and stent areas were smaller (BMS 7.2 ± 2.4 mm 2 , first-generation DES 6.1 ± 2.1 mm 2 , second-generation DES 5.7 ± 2.0 mm 2 , p <0.001). Stent fracture was seen only in DES (first-generation DES 7 [5.0%], second-generation DES 8 [7.4%], p = 0.13). In conclusion, restenotic first- and second-generation DES were characterized by less neointimal hyperplasia, smaller stent areas, longer stent lengths, and more stent fractures than restenotic BMS.
Reported causes of in-stent restenosis (ISR) after implantation of bare metal stents (BMS) or first- or second-generation drug-eluting stents (DES) include neointimal hyperplasia (NIH); stent under expansion; strut fracture, stent deformation, or other mechanical complications; and neoatherosclerosis. The bioactivity of DES reduces neointimal growth, ISR, and repeat revascularization compared with BMS; however, there are limited data comparing first- and second-generation DES with BMS. Although coronary angiography is commonly used to evaluate ISR lesions, it is limited in its ability to assess ISR in detail. Conversely, intravascular imaging provides detailed analyses of causes of ISR that are not possible with angiography alone. The purpose of this study was to use intravascular ultrasound (IVUS) studies to compare restenosis in BMS, first-generation DES, and second-generation DES.
Methods
We retrospectively analyzed 298 ISR lesions (in-stent minimum lumen area [MLA] <4 mm 2 ) in 298 patients who underwent IVUS follow-up from October 2009 to April 2014 at our institution (NewYork-Presbyterian Hospital, New York, New York). Stents were divided into 3 groups: BMS, first-generation DES (Cypher [Cordis Corp, Miami Lakes, Florida], TAXUS [Boston Scientific, Natick, Massachusetts], and Endeavor [Medtronic, Santa Rosa, California]), and second-generation DES (Promus [Boston Scientific], XIENCE [Abbott Vascular, Santa Clara, California], and Resolute [Medtronic]). We classified Endeavor as first-generation DES and Resolute as second-generation DES, although both are zotarolimus-eluting stents (ZES). This study was approved by the institutional review board, and written informed consent was obtained from all patients.
Renal insufficiency was defined as creatinine clearance <60 ml/min as calculated by the Cockcroft–Gault equation. Clinical presentations at the time of ISR were divided into unstable angina (no ST-segment elevation and no biomarker release, but clinical presentation as angina at rest, new-onset angina, or a crescendo pattern of anginal occurrence), ST-segment elevation myocardial infarction (STEMI), non–ST-segment elevation myocardial infarction (NSTEMI, no ST-segment elevation but with biomarker release), and stable angina.
Angiographic restenosis was classified in the worst view as: (1) focal ISR (type I, defined as ISR ≤10 mm in length); (2) diffuse ISR (type II, defined as ISR >10 mm length but within the stent); (3) proliferative (type III, defined as ISR >10 mm in length but extending beyond the stent edges); and (4) total occlusion (type IV). Focal ISR was subclassified into IA (articulation or gap), IB (margin), IC (focal body), and ID (multifocal).
IVUS imaging was performed after 0.1- to 0.2-mg intracoronary nitroglycerine using a commercially available IVUS system (iLab with 40-MHz Atlantis SR Pro catheters; Boston Scientific, Fremont, California or Revolution with 45-MHz catheters; Volcano, Rancho Cordova, California). Quantitative IVUS analysis was performed using computerized planimetry (echoPlaque; INDEC Medical Systems, Santa Clara, California). IVUS measurements included the cross-sectional areas (CSA) of the external elastic membrane, lumen, stent, and NIH (stent minus intrastent lumen). The proximal and distal reference sites were measured at the most normal-looking cross sections within 5 mm proximal and distal to the stent, but before any side branch. ISR patterns were classified as focal (body or margin), multifocal (body or margin), proliferative, or diffuse, as previously reported. Stent fracture was classified as (1) partial fracture, (2) complete fracture, (3) longitudinal fracture with overlap, and (4) deformation with distortion or shortening, as previously reported. Stent malapposition was defined as blood speckle behind stent struts not overlying a side branch. IVUS analysis was done by agreement of 2 independent cardiologists (KG and AM) who were blinded to clinical and angiographic data.
Statistical analyses were performed using SPSS version 22 (IBM, Armonk, New York). Categorical variables were summarized as frequencies. Continuous variables were presented as mean ± standard deviation. A p value of <0.05 was considered statistically significant. IVUS parameters were compared among the groups (BMS, first-generation DES, and second-generation DES) using analysis of variance or chi-square statistics, as appropriate. Post hoc analyses of continuous variables were done using Bonferroni corrections for multiple comparisons. Post hoc analyses of categorical variables were done using standardized residuals.
Results
Among the 298 ISR lesions, there were 52 BMS, 138 first-generation DES (73 Cypher, 52 TAXUS, and 13 Endeavor), and 108 second-generation DES (39 Promus, 66 XIENCE, and 3 Resolute). Baseline patient are summarized in Table 1 . Patient age was 66.6 ± 1.1 years, 74% were men, and 48% had diabetes mellitus. Hypertension was significantly more common in patients with DES than in patients with BMS. Very few of the current restenotic stents were originally implanted for STEMI or NSTEMI indications. The duration between stent implantation and BMS restenosis (70.0 ± 66.7 months) was longer than first-generation DES restenosis (40.1 ± 25.7 months) that was, in turn, longer than second-generation DES restenosis (13.1 ± 9.5 months).
| Variable | BMS (n = 52) | First-generation DES (n = 138) | Second-generation DES (n = 108) | p Value |
|---|---|---|---|---|
| Age (years) | 66.4±9.7 | 66.2±10.1 | 67.3±12.9 | 0.72 |
| Male sex | 39 (75%) | 106 (77%) | 76 (70%) | 0.52 |
| Diabetes | 19 (37%) | 67 (49%) | 58 (54%) | 0.13 |
| Insulin-treated | 6 (12%) | 12 (8.7%) | 19(18%) | 0.11 |
| Hyperlipidemia | 47 (90%) | 129 (94%) | 103 (95%) | 0.48 |
| Hypertension ∗ | 44 (85%) | 125 (91%) | 105 (97%) | 0.02 |
| Current smoker | 4 (7.7%) | 15 (11%) | 7 (6.5%) | 0.46 |
| Prior myocardial infarction | 21 (40%) | 45 (33%) | 41 (38%) | 0.52 |
| Prior coronary artery bypass grafting | 7 (14%) | 27 (20%) | 31 (29%) | 0.06 |
| Renal insufficiency † | 6 (12%) | 34 (25%) | 32 (30%) | 0.04 |
| Dialysis | 0 | 12 (8.7%) | 10 (9.3%) | 0.08 |
| Ejection fraction <40% | 7 (13.5%) | 14 (10%) | 10 (9.3%) | 0.71 |
| Medications | ||||
| Statin | 46 (89%) | 127 (92%) | 94 (87%) | 0.43 |
| Clopidogrel ‡ | 35 (67%) | 115 (83%) | 96 (89%) | 0.003 |
| Aspirin | 45 (87%) | 128 (93%) | 100 (93%) | 0.35 |
| Beta blocker | 43 (83%) | 118 (86%) | 92 (85%) | 0.89 |
| Angiotensin-converting enzyme inhibitor or angiotensin receptor blocker | 32 (62%) | 97 (70%) | 77 (71%) | 0.42 |
| Clinical presentation at the time of in-stent restentosis | ||||
| Unstable angina | 26 (50%) | 70 (51%) | 48 (44%) | 0.60 |
| ST-elevation myocardial infarction | 1 (1.9%) | 2 (1.4%) | 1 (0.9%) | 0.87 |
| Non–ST-elevation myocardial infarction | 1 (1.9%) | 9 (6.5%) | 7 (6.5%) | 0.43 |
| Stable angina | 24 (46%) | 57 (41%) | 52 (48%) | 0.55 |
| Duration between index and stent failure (months) § | 70.0±66.7 | 40.1±25.7 | 13.1±9.5 | <0.001 |
∗ BMS, p = 0.04, second-generation DES p=0.01.
‡ BMS, p = 0.001, second-generation DES, p = 0.03.
§ BMS versus first-generation DES, p <0.001, first-generation DES versus second-generation DES, p<0.001, BMS versus second-generation DES, p <0.001.
Lesion and procedural characteristics are provided in Table 2 . ISR lesions at the time of stent implantation were more common in the first- or second-generation restenotic DES versus restenotic BMS. Restenotic stents were longer in second-generation DES (32.1 ± 18.7 mm) compared to the first-generation DES (29.4 ± 16.1 mm) and BMS (21.8 ± 13.5 mm). Stent diameter was larger in BMS (3.27 ± 0.51 mm) compared with first-generation DES (2.95 ± 0.37 mm) and second-generation DES (3.00 ± 0.37 mm). There were no significant differences in angiographic or IVUS qualitative ISR classifications in the 3 groups either overall or looking at short and long stents separately (data not shown).
| Variable | BMS (n = 52) | First-generation DES (n = 138) | Second-generation DES (n = 108) | p Value |
|---|---|---|---|---|
| Target coronary vessel | ||||
| Left anterior descending ∗ | 29 (56%) | 66 (48%) | 38 (35%) | 0.03 |
| Left circumflex | 7 (14%) | 29 (21%) | 23 (21%) | 0.45 |
| Right | 15 (29%) | 37 (27%) | 42 (39%) | 0.12 |
| Left main | 1 (1.9%) | 5 (3.6%) | 4 (3.7%) | 0.82 |
| Vein graft | 0 | 1 (0.7%) | 1 (0.9%) | 0.79 |
| Lesion location | ||||
| Proximal † | 11 (21%) | 36 (26%) | 42 (39%) | 0.03 |
| Mid ‡ | 38 (73%) | 80 (58%) | 45 (42%) | 0.001 |
| Distal | 3 (5.8%) | 22 (16%) | 21 (19%) | 0.08 |
| Index lesion type | ||||
| In-stent restenosis § | 7 (14%) | 52 (38%) | 37 (34%) | 0.005 |
| Bifurcation | 15 (29%) | 43 (31%) | 29 (27%) | 0.76 |
| Ostial | 2 (3.8%) | 11 (8%) | 10 (9.3%) | 0.48 |
| Chronic total occlusion | 1 (1.9%) | 2 (1.4%) | 4 (3.7%) | 0.50 |
| Stent number per lesion ‖ | 1.3±0.6 | 1.6±0.8 | 1.8±0.8 | 0.001 |
| Total stent length (mm) ¶ | 21.8±13.5 | 29.4±16.1 | 32.1±18.7 | 0.003 |
| Average stent diameter (mm) # | 3.27±0.51 | 2.95±0.37 | 3.00±0.37 | <0.001 |
∗ Second-generation DES p = 0.012.
† second-generation DES, p = 0.009.
‡ BMS p = 0.004, second-generation DES p <0.001.
‖ BMS versus second-generation DES, p = 0.001.
¶ BMS versus first-generation DES, p = 0.04, BMS versus second-generation DES, p = 0.002.
# BMS versus first-generation DES, p <0.001, BMS versus second-generation DES, p = 0.001.
Although proximal and distal reference external elastic membrane and lumen areas and lesion site external elastic membrane areas were similar in the 3 groups, stent area at the MLA site was larger in BMS compared with the first- or second-generation DES ( Table 3 ); however, the frequency with which NIH occupied >50% of the stent area at the MLA site was similar in the 3 groups. These findings were supported by the analysis at the minimum stent area (MSA) site and at the maximum NIH site—in particular, the frequency with which the MSA measured <5 mm 2 . The MLA site was located at the maximum NIH site in 75% of BMS, 69% of first-generation DES, and 62% of second-generation DES (p = 0.24); conversely, the MLA site was located at the MSA site in 4% of BMS, 7% of first-generation DES, and 10% of second-generation DES (p = 0.36).
| Variable | BMS (n = 52) | First- Generation DES (n = 138) | Second- Generation DES (n = 108) | p Value |
|---|---|---|---|---|
| Proximal reference | ||||
| EEM CSA (mm 2 ) | 14.1±4.5 | 13.6±4.0 | 13.3±3.8 | 0.72 |
| Lumen CSA (mm 2 ) | 6.9±2.5 | 7.1±2.4 | 6.9±2.1 | 0.88 |
| Distal reference | ||||
| EEM CSA (mm 2 ) | 10.8±4.4 | 9.1±3.8 | 10.2±5.2 | 0.09 |
| Lumen CSA (mm 2 ) | 6.0±2.3 | 5.2±1.8 | 5.7±2.7 | 0.18 |
| MLA site | ||||
| EEM CSA (mm 2 ) | 14.2±3.8 | 15.1±7.2 | 13.5±4.4 | 0.15 |
| Minimum lumen area (mm 2 ) | 2.7±0.7 | 2.5±0.8 | 2.6±0.7 | 0.26 |
| Stent CSA (mm 2 ) ∗ | 7.2±2.4 | 6.1±2.1 | 5.7±2.0 | <0.001 |
| Stent CSA <5 mm 2 | 11 (22%) | 40 (31%) | 40 (40%) | 0.07 |
| Neointimal hyperplasia (%) † | 58.6±15.5 | 52.6±20.9 | 48.2±22.2 | 0.02 |
| Neointimal hyperplasia >50% | 40 (77%) | 104 (75%) | 71 (66%) | 0.25 |
| MSA site | ||||
| Minimum stent area (mm 2 ) ‡ | 6.4±2.2 | 4.9±1.6 | 4.7±1.6 | <0.001 |
| Minimum stent area <5 mm 2 § | 15 (29%) | 78 (57%) | 75 (69%) | <0.001 |
| Maximum neointimal hyperplasia site | ||||
| Neointimal hyperplasia (%) ‖ | 60.9±12.8 | 56.0±16.0 | 52.5±17.0 | 0.008 |
| Neointimal hyperplasia >50% | 42 (81%) | 107 (78%) | 74 (69%) | 0.15 |
∗ BMS versus first-generation DES, p = 0.004, BMS versus second-generation DES, p <0.001.
† BMS versus second-generation DES, p = 0.02.
‡ BMS versus first-generation DES, p <0.001, BMS versus second-generation DES, p <0.001.
§ BMS p <0.001, second-generation DES p <0.001.
Figure 1 shows the frequency distribution of the stent CSA at the MLA site and the percent NIH at the MLA site in the 3 groups. In Figure 2 , restenotic BMS and restenotic first- and second-generation DES were divided into dominant under expansion, dominant NIH, and mixed under expansion and NIH based on analysis at the MLA site.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
